DETAILED ACTION
Response to Arguments
The amendments filed 8/22/2025 have been entered and made of record.
Applicant's amendments and arguments filed 8/22/2025 have been considered but are moot because the independent claim 1 has been substantially amended with newly added limitation, and Applicant's arguments in view of the amendments filed 8/22/2025 have been fully considered but they are not persuasive:
First, Applicant states that the cited references, particularly Chavez Badiola as modified by SHAFIEE do not disclose two separated models: a) a neural network to generate a first clinical parameter from {the image data} of processing of the image of the embryo; and b) a second model {a predictive model} that generate a second clinical parameter from a set of classification parameters that not derived from imaging of the embryo;
However, the Examiner disagrees, because:
SHAFIEE clearly disclose two separated models: a) a neural network to generate a first clinical parameter from {the image data} of processing of the image of the embryo (see SHAFIEE: e.g., -- the convolutional neural network 104 is trained on a plurality of images of embryos taken on a third day of embryo development…--, in [0023]-[0025]); and
b) a second model {a predictive model} that generate a second clinical parameter from a set of classification parameters that not derived from imaging of the embryo (see -- another expert system (not shown). In practice, any of a variety of experts systems can be utilized in combination with the convolutional neural network, including support vector machines, random forest, self-organized maps, fuzzy logic systems, data fusion processes, ensemble methods, rule based systems, genetic algorithms, and artificial neural networks. It will be appreciated that the additional expert system may be trained on features from multiple stages of embryonic development as well as with features that are external to the images, such as biometric parameters of an egg donor, a sperm donor, or a recipient of the embryo.--, in [0026]);
apparently, “the expert system” is a second model, other than the first model of “neural network 104”;
and, claim 1 just limits a step of “generating a composite parameter….”, but does not further define, how to use, and what’s the processing functions or algorithms would further apply on the such generated “ composite parameter”, neither in any dependent claims 2-7;
In claim 8, claim 8 limits “an arbitrator configured to generate a composite parameter…”, but “herein arbitrator” is just a portion of “machine executable instructions”, however claim 8 still does not further define, how to use, and what’s the processing functions or algorithms would further apply on the such generated “ composite parameter”;
and, although dependent claim 12 limits “the arbitrator being configured to generate the composite parameter according to a majority vote among at least the first clinical parameter, the second clinical parameter, and the third clinical parameter”, which limits “a majority vote”, however does not limit the criteria, or what’s attributes or characteristics are being voting based on; and still does not further define, how to use, and what’s the processing functions or algorithms would further apply on the such generated “ composite parameter”;
Therefore, claims 1-20 are still not patentably distinguishable over the prior art reference(s). Further discussions are addressed in the prior art rejection section below.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Chavez Badiola (US 20220392062 A1, Date Filed: 2019-12-20), in view of SHAFIEE (WO 2019068073 A1, Date Published: 2019-04-04 {over one-year grace period}).
Re Claim 1, Chavez Badiola discloses a method for fully automated screening for aneuploidy in a human embryo (see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]), comprising:
obtaining an image of the embryo at an associated imager (see Chavez Badiola: e.g., -- an apparatus includes a classification module configured to determine classifiers to images of one to more cells to determine, for each image, a classification probability associated with each classifier.--, in [0003]-[0010]);
providing the image of the embryo to a neural network to generate a first clinical parameter (see Chavez Badiola: e.g., -- an apparatus includes a classification module configured to determine classifiers to images of one to more cells to determine, for each image, a classification probability associated with each classifier.--, in [0003]-[0010]; and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]);
retrieving a set of at least one parameter representing one of a patient receiving the embryo, an egg utilized to produce the human embryo (see Chavez Badiola: e.g., -- Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective),--, in [0060]-[0061]),
Chavez Badiola however does not explicitly disclose parameter representing a sperm used to create the embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg,
SHAFIEE discloses retrieving parameter representing a sperm used to create the embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg (see SHAFIEE: e.g., -- biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo.--, in [0027], and, --[0032] The system 300 includes an imager 306 that acquires an image of for each embryo on at least one selected day of development. This image is then provided to the convolutional neural network 302. It will be appreciated, however, that multiple images, taken from different days, can be used at one or multiple convolutional neural networks. The convolutional neural network 302 generates a plurality of values representing the morphology of the embryo from the provided one or more images. This plurality of values, as well as a plurality of features representing biometric parameters of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg to the expert system, are provided to the expert system 304. The biometric parameters can include, for example, an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. In the illustrated implementation, the biometric parameters are retrieved from an electronic health records (EHR) database 308.--, in [0032]);
Chavez Badiola and SHAFIEE are combinable as they are in the same field of endeavor: using neural network in analysis of embryo image. Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Chavez Badiola’s method using SHAFIEE’s teachings by including parameter representing a sperm used to create the embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg to Chavez Badiola’s a set of at least one parameter in order to use biometric parameters in determining a quality of the embryo (see SHAFIEE: e.g., [0027], and [0032]);
Chavez Badiola as modified by SHAFIEE further disclose generating a second clinical parameter a set of classification parameters including the set of at least one parameter at a predictive model, wherein none of the set of classification parameters are derived from imaging of the embryo (see Chavez Badiola: e.g., -- defined mathematical models that allow the identification of intensity patterns in 2, 3 or 4 dimensions such as roughness, contrast, brightness, saturation, smoothness, or particular shapes, for the prediction of embryo characteristics such as degree of collapse, or degradation, as well as the stage of the embryo; said parameters may be based on filters that identify a plurality of textures and/or other metrics based on segmentation of cell types: in which said plurality of textures may be at least one of those where a combination of texture detector masks are used such as 2-dimensional LAWS (Preferably 25, wherein the LAWS energy is obtained by detecting textures in embryo images which may be the standardized, or enhanced through machine vision strategies such as energy filters, Gaussian, Laplacian and edge detectors) to identify textures on the original image and at least one variant of the original. To generate the variants, entropy filters with different radius of influence and Gaussian blur are used; once this activity is completed, the automatic cropping technique is implemented, that is, the calculated textures are used, and the k-means algorithm is used with a k value of at least 2 to identify the pixels that belong to the background from those that belong to the embryo or any instrument or material present in the photograph that is not the embryo. Based on this 2, 3, or 4-dimensional mask, the edges are detected to crop the image containing the embryo. As an alternative to the automatic cropping based on the k-means algorithm, it is also possible to use an artificial intelligence-based object identifier, to identify and subtract from the image instruments, letters, or other artifacts unrelated to the embryo. To identify the stages of embryo development, a deep convolutional neural network model that can classify embryos into one of three stages is used: a) expanding, b) hatching and c) hatched; or it can function as a regressor defining the percentage that is inside the zona pellucida and the percentage that is outside. With the classifier technique, a probability value [0-1] is obtained, and the image corresponds to one of these three classes with an accuracy depending on the model used. To identify embryos that are collapsed (a natural process of embryos), a previously trained classifier can be used, which is identified with a probability index, or through a regressor that allows the identification of the percentage of collapse shown by the image of the embryo; to identify the degree of degradation of an embryo, one can use a pre-trained classifier that uses a probability index that a given image is in the “degraded’ or “normal” class, or a pre-trained regressor that identifies the percentage of degradation of the embryo through the image. To identify the degree of development of an embryo within a developmental curve, a previously trained classifier can be used, which uses a probability index indicating that a given image falls within one or several classes associated with the degree of evolution or growth of the embryo according to the expected growth given the embryo's conditions; as an alternative to the classifier, a previously trained regressor can be used to identify the percentile in which the embryo is located according to its growth and development, based on healthy embryos, or alternatively, statistical data can be used to locate the image of the embryo within a distribution of statistical parameters of growth such as the size of the different zones of the embryo. Therefore, at the end of this stage, a computer-implemented algorithm is available to identify the pixels or voxels that belong to at least one of the following five areas: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; the supervised training is performed through manual labels on the texture vectors for each pixel or voxel; where, from the predictions made at the pixel or voxel level, the predictions are subjected to a process to generate more homogeneous areas for each label, which involves the extraction of the blobs of a k-means (k=20) and a process of erosion and dilation of the zones. An alternative to identify the different zones is to use a neural network model containing an encoder and a decoder that associates each pixel or voxel to one of five labels: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis--, in [0060]-[0062]; also see SHAFIEE: e.g., -- the convolutional neural network 104 is trained on a plurality of images of embryos taken on a third day of embryo development…--, in [0023]-[0025]; and
{for b) a second model {a predictive model} that generate a second clinical parameter from a set of classification parameters that not derived from imaging of the embryo}; see SHAFIEE: e.g., -- another expert system (not shown). In practice, any of a variety of experts systems can be utilized in combination with the convolutional neural network, including support vector machines, random forest, self-organized maps, fuzzy logic systems, data fusion processes, ensemble methods, rule based systems, genetic algorithms, and artificial neural networks. It will be appreciated that the additional expert system may be trained on features from multiple stages of embryonic development as well as with features that are external to the images, such as biometric parameters of an egg donor, a sperm donor, or a recipient of the embryo.--, in [0026]; and, -- the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]);
generating a composite parameter, representing a likelihood of aneuploidy in the embryo, from the first clinical parameter and the second clinical parameter (see Chavez Badiola: e.g., -- defined mathematical models that allow the identification of intensity patterns in 2, 3 or 4 dimensions such as roughness, contrast, brightness, saturation, smoothness, or particular shapes, for the prediction of embryo characteristics such as degree of collapse, or degradation, as well as the stage of the embryo; said parameters may be based on filters that identify a plurality of textures and/or other metrics based on segmentation of cell types: in which said plurality of textures may be at least one of those where a combination of texture detector masks are used such as 2-dimensional LAWS (Preferably 25, wherein the LAWS energy is obtained by detecting textures in embryo images which may be the standardized, or enhanced through machine vision strategies such as energy filters, Gaussian, Laplacian and edge detectors) to identify textures on the original image and at least one variant of the original. To generate the variants, entropy filters with different radius of influence and Gaussian blur are used; once this activity is completed, the automatic cropping technique is implemented, that is, the calculated textures are used, and the k-means algorithm is used with a k value of at least 2 to identify the pixels that belong to the background from those that belong to the embryo or any instrument or material present in the photograph that is not the embryo. Based on this 2, 3, or 4-dimensional mask, the edges are detected to crop the image containing the embryo. As an alternative to the automatic cropping based on the k-means algorithm, it is also possible to use an artificial intelligence-based object identifier, to identify and subtract from the image instruments, letters, or other artifacts unrelated to the embryo. To identify the stages of embryo development, a deep convolutional neural network model that can classify embryos into one of three stages is used: a) expanding, b) hatching and c) hatched; or it can function as a regressor defining the percentage that is inside the zona pellucida and the percentage that is outside. With the classifier technique, a probability value [0-1] is obtained, and the image corresponds to one of these three classes with an accuracy depending on the model used. To identify embryos that are collapsed (a natural process of embryos), a previously trained classifier can be used, which is identified with a probability index, or through a regressor that allows the identification of the percentage of collapse shown by the image of the embryo; to identify the degree of degradation of an embryo, one can use a pre-trained classifier that uses a probability index that a given image is in the “degraded’ or “normal” class, or a pre-trained regressor that identifies the percentage of degradation of the embryo through the image. To identify the degree of development of an embryo within a developmental curve, a previously trained classifier can be used, which uses a probability index indicating that a given image falls within one or several classes associated with the degree of evolution or growth of the embryo according to the expected growth given the embryo's conditions; as an alternative to the classifier, a previously trained regressor can be used to identify the percentile in which the embryo is located according to its growth and development, based on healthy embryos, or alternatively, statistical data can be used to locate the image of the embryo within a distribution of statistical parameters of growth such as the size of the different zones of the embryo. Therefore, at the end of this stage, a computer-implemented algorithm is available to identify the pixels or voxels that belong to at least one of the following five areas: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; the supervised training is performed through manual labels on the texture vectors for each pixel or voxel; where, from the predictions made at the pixel or voxel level, the predictions are subjected to a process to generate more homogeneous areas for each label, which involves the extraction of the blobs of a k-means (k=20) and a process of erosion and dilation of the zones. An alternative to identify the different zones is to use a neural network model containing an encoder and a decoder that associates each pixel or voxel to one of five labels: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]; also see SHAFIEE: e.g., -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; and, --the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]).
Re Claim 2, Chavez Badiola as modified by SHAFIEE further disclose wherein providing the image of the embryo to the neural network comprises providing the image of the embryo to a recurrent neural network (see SHAFIEE: e.g., -- In one implementation, the neural network 104 can be a convolutional neural network, which is a feed-forward artificial neural network that includes convolutional layers, which effectively apply a convolution to the values at the preceding layer of the network to emphasize various sets of features within an image. In a convolutional layer, each neuron is connected only to a proper subset of the neurons in the preceding layer, referred to as the receptive field of the neuron. In the illustrated example, the convolutional neural network is implemented using the Xception architecture. In one implementation, at least one chromatic value {e.g., a value for an RGB color channel, a YCrCb color channel, or a grayscale brightness) associated with each pixel is provided as an initial input to the convolutional neural network. [0020] In another implementation, the neural network 104 can be implemented as a recurrent neural network. In a recurrent neural network, the connections between nodes in the network are selected to form a directed graph along a sequence, allowing it to exhibit dynamic temporal behavior. In another implementation, the neural network 104 is implemented and trained as a discriminative network in a generative adversarial model, in which a generative neural network and the discriminative network provide mutual feedback to one another, such that the generative neural network produces increasingly sophisticated samples for the discriminative network to attempt to classify.--, in [0019]-[0020]).
Re Claim 3, Chavez Badiola as modified by SHAFIEE further disclose wherein providing the image of the embryo to the neural network comprises providing the image of the embryo to a discriminative classifier trained as part of a generative adversarial network (see SHAFIEE: e.g., -- [0027] FIG. 2 provides one example of a system 200 in which a convolutional neural network 202 is used in combination with another expert system 204. In this instance, the other expert system is a genetic algorithm 204 that receives the output of the convolutional neural network 202 and generates a metric representing the quality of an embryo being evaluated. Each of the convolutional neural network 202 and the genetic algorithm 204 can be implemented as software instructions stored on a non-transitory computer readable medium and executed by an associated processor The quality of the embryo can be a binary categorical variable, representing the success or failure of the implantation or a numerical grade of the embryo.. In addition to the output of the convolutional neural network, the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]).
Re Claim 4, Chavez Badiola as modified by SHAFIEE further disclose wherein providing the image of the embryo to the neural network comprises providing the image of the embryo to a convolutional neural network (see Chavez Badiola: e.g., -- defined mathematical models that allow the identification of intensity patterns in 2, 3 or 4 dimensions such as roughness, contrast, brightness, saturation, smoothness, or particular shapes, for the prediction of embryo characteristics such as degree of collapse, or degradation, as well as the stage of the embryo; said parameters may be based on filters that identify a plurality of textures and/or other metrics based on segmentation of cell types: in which said plurality of textures may be at least one of those where a combination of texture detector masks are used such as 2-dimensional LAWS (Preferably 25, wherein the LAWS energy is obtained by detecting textures in embryo images which may be the standardized, or enhanced through machine vision strategies such as energy filters, Gaussian, Laplacian and edge detectors) to identify textures on the original image and at least one variant of the original. To generate the variants, entropy filters with different radius of influence and Gaussian blur are used; once this activity is completed, the automatic cropping technique is implemented, that is, the calculated textures are used, and the k-means algorithm is used with a k value of at least 2 to identify the pixels that belong to the background from those that belong to the embryo or any instrument or material present in the photograph that is not the embryo. Based on this 2, 3, or 4-dimensional mask, the edges are detected to crop the image containing the embryo. As an alternative to the automatic cropping based on the k-means algorithm, it is also possible to use an artificial intelligence-based object identifier, to identify and subtract from the image instruments, letters, or other artifacts unrelated to the embryo. To identify the stages of embryo development, a deep convolutional neural network model that can classify embryos into one of three stages is used: a) expanding, b) hatching and c) hatched; or it can function as a regressor defining the percentage that is inside the zona pellucida and the percentage that is outside. With the classifier technique, a probability value [0-1] is obtained, and the image corresponds to one of these three classes with an accuracy depending on the model used. To identify embryos that are collapsed (a natural process of embryos), a previously trained classifier can be used, which is identified with a probability index, or through a regressor that allows the identification of the percentage of collapse shown by the image of the embryo; to identify the degree of degradation of an embryo, one can use a pre-trained classifier that uses a probability index that a given image is in the “degraded’ or “normal” class, or a pre-trained regressor that identifies the percentage of degradation of the embryo through the image. To identify the degree of development of an embryo within a developmental curve, a previously trained classifier can be used, which uses a probability index indicating that a given image falls within one or several classes associated with the degree of evolution or growth of the embryo according to the expected growth given the embryo's conditions; as an alternative to the classifier, a previously trained regressor can be used to identify the percentile in which the embryo is located according to its growth and development, based on healthy embryos, or alternatively, statistical data can be used to locate the image of the embryo within a distribution of statistical parameters of growth such as the size of the different zones of the embryo. Therefore, at the end of this stage, a computer-implemented algorithm is available to identify the pixels or voxels that belong to at least one of the following five areas: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; the supervised training is performed through manual labels on the texture vectors for each pixel or voxel; where, from the predictions made at the pixel or voxel level, the predictions are subjected to a process to generate more homogeneous areas for each label, which involves the extraction of the blobs of a k-means (k=20) and a process of erosion and dilation of the zones. An alternative to identify the different zones is to use a neural network model containing an encoder and a decoder that associates each pixel or voxel to one of five labels: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis--, in [0060]-[0062]; also see SHAFIEE: e.g., -- [0027] FIG. 2 provides one example of a system 200 in which a convolutional neural network 202 is used in combination with another expert system 204. In this instance, the other expert system is a genetic algorithm 204 that receives the output of the convolutional neural network 202 and generates a metric representing the quality of an embryo being evaluated. Each of the convolutional neural network 202 and the genetic algorithm 204 can be implemented as software instructions stored on a non-transitory computer readable medium and executed by an associated processor The quality of the embryo can be a binary categorical variable, representing the success or failure of the implantation or a numerical grade of the embryo.. In addition to the output of the convolutional neural network, the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]).
Re Claim 5, Chavez Badiola as modified by SHAFIEE further disclose wherein the set of at least one parameter includes an age of the egg donor who provided the oocyte fertilized to produce the embryo (see SHAFIEE: e.g., -- The convolutional neural network 302 generates a plurality of values representing the morphology of the embryo from the provided one or more images. This plurality of values, as well as a plurality of features representing biometric parameters of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg to the expert system, are provided to the expert system 304. The biometric parameters can include, for example, an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. In the illustrated implementation, the biometric parameters are retrieved from an electronic health records (EHR) database 308.--, in [0032]).
Re Claim 6, Chavez Badiola as modified by SHAFIEE further disclose wherein the set of at least one parameter includes a number of embryos that were normally fertilized from oocytes harvested with the oocyte fertilized to produce the embryo (see SHAFIEE: e.g., -- the neural network 104 is trained on a plurality of images of embryos taken on a first day of embryo development, for example, at eighteen hours after fertilization, that are classified into either a first class, representing normal fertilization of the embryo, or a second class, representing an abnormal fertilization of the embryo. For the purpose of this application, a normally fertilized embryo is an embryo that contains two pronuclei and an abnormally fertilized embryo is an embryo with any other number of pronuclei. Accordingly, in this implementation, abnormal embryos can be detected at an early stage without the intervention of a trained embryologist.--, in [0022]; also see Chavez Badiola: e.g., -- [0006] each first classification is associated with a first distinct number of cells and determines the first classification probability for each image based on a plurality of cell features that include one or more machine-learned cell features; and [0007] the first probability of classification indicates a first estimated likelihood that the first distinct number of cells associated with each first classifier is shown in each image, and each of the plurality of images thereby has a plurality of the first classification probabilities associated therewith; and [0008] classifying each image as showing a second number of cells based on the distinct first number of cells associated with each first classifier and the plurality of first classification probabilities associated with it. [0009] Each classifier is associated with a distinct first number of cells and is designed to determine the classification probability of each image based on cell characteristics, including one or more machine-learned cell features. The classification probability indicates an estimated probability that the first distinct number of cells will be displayed in each image. Therefore, each of the images has classification probabilities associated.--, in [0006]-[0010], and [0015]).
Re Claim 7, Chavez Badiola as modified by SHAFIEE further disclose the set of at least one parameter includes a value representing a quality of the sperm used to fertilize the embryo (see SHAFIEE: e.g., -- [0027] FIG. 2 provides one example of a system 200 in which a convolutional neural network 202 is used in combination with another expert system 204. In this instance, the other expert system is a genetic algorithm 204 that receives the output of the convolutional neural network 202 and generates a metric representing the quality of an embryo being evaluated. Each of the convolutional neural network 202 and the genetic algorithm 204 can be implemented as software instructions stored on a non-transitory computer readable medium and executed by an associated processor The quality of the embryo can be a binary categorical variable, representing the success or failure of the implantation or a numerical grade of the embryo.. In addition to the output of the convolutional neural network, the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]).
Re Claims 8, claim 8 is the corresponding system claim to claim 1, respectively. Claim 8 is rejected for the similar reasons for claim 1. See above discussions with regard to claim 1 respectively. Chavez Badiola as modified by SHAFIEE further disclose A system for fully automated screening for aneuploidy in a human embryo, the system comprising: a processor; and a non-transitory computer readable medium storing instructions executable by the processor to perform the method (see SHAFIEE: e.g., --[0039] The system 500 can includes a system bus 502, a processing unit 504, a system memory 506, memory devices 508 and 510, a communication interface 512 (e.g., a network interface), a communication link 514, a display 516 (e.g., a video screen), and an input device 518 (e.g., a keyboard and/or a mouse). The system bus 502 can be in communication with the processing unit 504 and the system memory 506. The additional memory devices 508 and 510, such as a hard disk drive, server, stand-alone database, or other non-volatile memory, can also be in communication with the system bus 502. The system bus 502 interconnects the processing unit 504, the memory devices 506-510, the communication interface 512, the display 516, and the input device 518. In some examples, the system bus 502 also interconnects an additional port (not shown), such as a universal serial bus (USB) port. [0040] The system 500 could be implemented in a computing cloud. In such a situation, features of the system 500, such as the processing unit 504, the communication interface 512, and the memory devices 508 and 510 could be representative of a single instance of hardware or multiple instances of hardware with applications executing across the multiple of instances (i.e., distributed) of hardware (e.g., computers, routers, memory, processors, or a combination thereof).--, in [0039]-[0044]; also see Chavez Badiola: e.g., Fig. 7, and, -- an apparatus includes a classification module configured to determine classifiers to images of one to more cells to determine, for each image, a classification probability associated with each classifier.--, in [0003]-[0010]; and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0060]-[0062], and [0094]).
Re Claim 9, Chavez Badiola as modified by SHAFIEE further disclose a feature extractor that generates a feature vector representing the image and provides the feature vector to the predictive model, the a predictive model configured to generate the second clinical parameter from at least one parameter representing the set of at least one parameter (see Chavez Badiola: e.g., --[0036] Measures the properties or characteristics of the entire blastocyst, including zona pellucida, trophectoderm, and inner cell mass. [0037] It extracts characteristics by identifying different cell types, mainly structures and patterns of the blastocyst, without extracting characteristics of the first cell divisions and their behavior over time, that is to say when there is still no cell differentiation identifiable with conventional instruments (microscopes and specialized cameras). [0038] It predicts the prognosis of a pregnancy and/or ploidy (in other words, the result of a genetic study and implantation success). [0039] Embryos are used from day 5 onwards (blastocyst), and the features obtained are based on structures or patterns, which are applied to the whole embryo and are not performed for each type of cell. [0040] Artificial intelligence and computer vision algorithms are used to extract important or predominant features of the embryo (features) and train the model. [0041] OBJECTIVE: To compare different embryos, based on the extracted features, and create a ranking to determine, according to their ploidy and/or implantation potential, which embryo has the highest potential to produce a pregnancy.--, in [0036]-[0041]).
Re Claim 10, Chavez Badiola as modified by SHAFIEE further disclose wherein the arbitrator is configured to generate the composite parameter as a weighted linear combination of at least the first clinical parameter and the second clinical parameter (see SHAFIEE: e.g., -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]).
Re Claim 11, Chavez Badiola as modified by SHAFIEE further disclose wherein the arbitrator is configured to generate the composite parameter as an average of at least the first clinical parameter and the second clinical parameter (see SHAFIEE: e.g., -- [0029] A second training set of images, representing embryos of known quality, can be generated at the imager 206 and provided to the trained convolutional neural network 202 to generate a set of convolutional neural network outputs and known embryo qualities. For example, the known quality can be an outcome of the implantation or an average {e.g., arithmetic mean or median) of numerical grades provided via manual assessment by a set of embryologists from different institutions. This set of outputs and the embryo qualities can be provided as training data to the genetic algorithm.--, in [0029]; also see: -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]).
Re Claim 12, Chavez Badiola as modified by SHAFIEE further disclose the predictive model is a first predictive model, the set of at least one parameter is a first set of at least one parameter, and the system further comprises a second predictive model configured to generate a third clinical parameter from a second set of at least one parameter, each of the first clinical parameter, the second clinical parameter, and the third clinical parameter being a categorical parameter and the arbitrator being configured to generate the composite parameter according to a majority vote among at least the first clinical parameter, the second clinical parameter, and the third clinical parameter (see SHAFIEE: e.g., -- [0029] A second training set of images, representing embryos of known quality, can be generated at the imager 206 and provided to the trained convolutional neural network 202 to generate a set of convolutional neural network outputs and known embryo qualities. For example, the known quality can be an outcome of the implantation or an average {e.g., arithmetic mean or median) of numerical grades provided via manual assessment by a set of embryologists from different institutions. This set of outputs and the embryo qualities can be provided as training data to the genetic algorithm.--, in [0029]; also see: -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]; also see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]; and, [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]).
Re Claim 13, Chavez Badiola as modified by SHAFIEE further disclose wherein the predicted model is implemented as a support vector machine (see SHAFIEE: e.g., -- [0029] A second training set of images, representing embryos of known quality, can be generated at the imager 206 and provided to the trained convolutional neural network 202 to generate a set of convolutional neural network outputs and known embryo qualities. For example, the known quality can be an outcome of the implantation or an average {e.g., arithmetic mean or median) of numerical grades provided via manual assessment by a set of embryologists from different institutions. This set of outputs and the embryo qualities can be provided as training data to the genetic algorithm.--, in [0029]; also see: -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]; also see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]; and, [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]).
Re Claim 14, Chavez Badiola as modified by SHAFIEE further disclose wherein the predicted model is implemented as a fully- connected feed-forward neural network (see Chavez Badiola: e.g., -- provides the set of pre-processed images of the embryo, classified, and ranked to an expert system characterized in that the expert system is a feedforward neural network.--, in claim 6; also see: -- [0077] The proposed system then uses the previously described list of characteristics and feeds them into a previously trained AI model to predict the prognosis of each embryo. This resulted in a list of probability values that each list of characteristics associated with each embryo image belongs to the good prognosis class. The embryos are ranked in descending order according to their probability value of belonging to the good prognosis class using letters of the alphabet in order, therefore the letter ‘A’ is assigned to the embryo with the best prognosis.--, in [0077]).
Re Claim 15, Chavez Badiola as modified by SHAFIEE further disclose wherein the set of at least one parameter includes at least one of an age of the egg, an age of an egg donor, an age of a sperm donor, a method of fertilization for the embryo, a hormonal profile of the egg donor, a past diagnosis of a condition of the egg donor, and a past diagnosis of a condition of the sperm donor (see SHAFIEE: e.g., -- biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo.--, in [0027], and, --[0032] The system 300 includes an imager 306 that acquires an image of for each embryo on at least one selected day of development. This image is then provided to the convolutional neural network 302. It will be appreciated, however, that multiple images, taken from different days, can be used at one or multiple convolutional neural networks. The convolutional neural network 302 generates a plurality of values representing the morphology of the embryo from the provided one or more images. This plurality of values, as well as a plurality of features representing biometric parameters of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg to the expert system, are provided to the expert system 304. The biometric parameters can include, for example, an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. In the illustrated implementation, the biometric parameters are retrieved from an electronic health records (EHR) database 308.--, in [0032]).
Re Claim 16, Chavez Badiola discloses a method for fully automated screening for aneuploidy in a human embryo (see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]), comprising:
obtaining an image of the embryo at an associated imager (see Chavez Badiola: e.g., -- an apparatus includes a classification module configured to determine classifiers to images of one to more cells to determine, for each image, a classification probability associated with each classifier.--, in [0003]-[0010]);
generating a first clinical parameter from the image of the embryo at a convolutional neural network (see Chavez Badiola: e.g., -- an apparatus includes a classification module configured to determine classifiers to images of one to more cells to determine, for each image, a classification probability associated with each classifier.--, in [0003]-[0010]; and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]);
retrieving a set of parameters from an associated memory (see Chavez Badiola: e.g., -- Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective),--, in [0060]-[0061]),
Chavez Badiola however does not explicitly disclose the set of parameters comprising an age of the egg donor who provided the oocyte fertilized to produce the embryo, a value representing a quality of the sperm used to create the embryo, and a number of embryos that were normally fertilized from oocytes harvested with the oocyte fertilized to produce the embryo;
SHAFIEE discloses the set of parameters comprising an age of the egg donor who provided the oocyte fertilized to produce the embryo, a value representing a quality of the sperm used to create the embryo, and a number of embryos that were normally fertilized from oocytes harvested with the oocyte fertilized to produce the embryo (see SHAFIEE: e.g., -- the neural network 104 is trained on a plurality of images of embryos taken on a first day of embryo development, for example, at eighteen hours after fertilization, that are classified into either a first class, representing normal fertilization of the embryo, or a second class, representing an abnormal fertilization of the embryo. For the purpose of this application, a normally fertilized embryo is an embryo that contains two pronuclei and an abnormally fertilized embryo is an embryo with any other number of pronuclei. Accordingly, in this implementation, abnormal embryos can be detected at an early stage without the intervention of a trained embryologist.--, in [0022]; and, -- biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo.--, in [0027], and, --[0032] The system 300 includes an imager 306 that acquires an image of for each embryo on at least one selected day of development. This image is then provided to the convolutional neural network 302. It will be appreciated, however, that multiple images, taken from different days, can be used at one or multiple convolutional neural networks. The convolutional neural network 302 generates a plurality of values representing the morphology of the embryo from the provided one or more images. This plurality of values, as well as a plurality of features representing biometric parameters of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg to the expert system, are provided to the expert system 304. The biometric parameters can include, for example, an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. In the illustrated implementation, the biometric parameters are retrieved from an electronic health records (EHR) database 308.--, in [0032]);
Chavez Badiola and SHAFIEE are combinable as they are in the same field of endeavor: using neural network in analysis of embryo image. Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Chavez Badiola’s method using SHAFIEE’s teachings by including the set of parameters comprising an age of the egg donor who provided the oocyte fertilized to produce the embryo, a value representing a quality of the sperm used to create the embryo, and a number of embryos that were normally fertilized from oocytes harvested with the oocyte fertilized to produce the embryo to Chavez Badiola’s a set of at least one parameter in order to use biometric parameters in determining a quality of the embryo (see SHAFIEE: e.g., [0022], [0027], and [0032]);
Chavez Badiola as modified by SHAFIEE further disclose generating a second clinical parameter at a first predictive model from a set of classification parameters including first subset of the set of parameters, wherein none of the set of classification parameters are derived from imaging of the embryo (see Chavez Badiola: e.g., -- defined mathematical models that allow the identification of intensity patterns in 2, 3 or 4 dimensions such as roughness, contrast, brightness, saturation, smoothness, or particular shapes, for the prediction of embryo characteristics such as degree of collapse, or degradation, as well as the stage of the embryo; said parameters may be based on filters that identify a plurality of textures and/or other metrics based on segmentation of cell types: in which said plurality of textures may be at least one of those where a combination of texture detector masks are used such as 2-dimensional LAWS (Preferably 25, wherein the LAWS energy is obtained by detecting textures in embryo images which may be the standardized, or enhanced through machine vision strategies such as energy filters, Gaussian, Laplacian and edge detectors) to identify textures on the original image and at least one variant of the original. To generate the variants, entropy filters with different radius of influence and Gaussian blur are used; once this activity is completed, the automatic cropping technique is implemented, that is, the calculated textures are used, and the k-means algorithm is used with a k value of at least 2 to identify the pixels that belong to the background from those that belong to the embryo or any instrument or material present in the photograph that is not the embryo. Based on this 2, 3, or 4-dimensional mask, the edges are detected to crop the image containing the embryo. As an alternative to the automatic cropping based on the k-means algorithm, it is also possible to use an artificial intelligence-based object identifier, to identify and subtract from the image instruments, letters, or other artifacts unrelated to the embryo. To identify the stages of embryo development, a deep convolutional neural network model that can classify embryos into one of three stages is used: a) expanding, b) hatching and c) hatched; or it can function as a regressor defining the percentage that is inside the zona pellucida and the percentage that is outside. With the classifier technique, a probability value [0-1] is obtained, and the image corresponds to one of these three classes with an accuracy depending on the model used. To identify embryos that are collapsed (a natural process of embryos), a previously trained classifier can be used, which is identified with a probability index, or through a regressor that allows the identification of the percentage of collapse shown by the image of the embryo; to identify the degree of degradation of an embryo, one can use a pre-trained classifier that uses a probability index that a given image is in the “degraded’ or “normal” class, or a pre-trained regressor that identifies the percentage of degradation of the embryo through the image. To identify the degree of development of an embryo within a developmental curve, a previously trained classifier can be used, which uses a probability index indicating that a given image falls within one or several classes associated with the degree of evolution or growth of the embryo according to the expected growth given the embryo's conditions; as an alternative to the classifier, a previously trained regressor can be used to identify the percentile in which the embryo is located according to its growth and development, based on healthy embryos, or alternatively, statistical data can be used to locate the image of the embryo within a distribution of statistical parameters of growth such as the size of the different zones of the embryo. Therefore, at the end of this stage, a computer-implemented algorithm is available to identify the pixels or voxels that belong to at least one of the following five areas: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; the supervised training is performed through manual labels on the texture vectors for each pixel or voxel; where, from the predictions made at the pixel or voxel level, the predictions are subjected to a process to generate more homogeneous areas for each label, which involves the extraction of the blobs of a k-means (k=20) and a process of erosion and dilation of the zones. An alternative to identify the different zones is to use a neural network model containing an encoder and a decoder that associates each pixel or voxel to one of five labels: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis--, in [0060]-[0062]; also see SHAFIEE: e.g., -- the convolutional neural network 104 is trained on a plurality of images of embryos taken on a third day of embryo development…--, in [0023]-[0025]; and
{for b) a second model {a predictive model} that generate a second clinical parameter from a set of classification parameters that not derived from imaging of the embryo}; see SHAFIEE: e.g., -- another expert system (not shown). In practice, any of a variety of experts systems can be utilized in combination with the convolutional neural network, including support vector machines, random forest, self-organized maps, fuzzy logic systems, data fusion processes, ensemble methods, rule based systems, genetic algorithms, and artificial neural networks. It will be appreciated that the additional expert system may be trained on features from multiple stages of embryonic development as well as with features that are external to the images, such as biometric parameters of an egg donor, a sperm donor, or a recipient of the embryo.--, in [0026]; and, -- the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]);
generating a third clinical parameter at a second predictive model from a second subset of the set of parameters (see Chavez Badiola: e.g., -- defined mathematical models that allow the identification of intensity patterns in 2, 3 or 4 dimensions such as roughness, contrast, brightness, saturation, smoothness, or particular shapes, for the prediction of embryo characteristics such as degree of collapse, or degradation, as well as the stage of the embryo; said parameters may be based on filters that identify a plurality of textures and/or other metrics based on segmentation of cell types: in which said plurality of textures may be at least one of those where a combination of texture detector masks are used such as 2-dimensional LAWS (Preferably 25, wherein the LAWS energy is obtained by detecting textures in embryo images which may be the standardized, or enhanced through machine vision strategies such as energy filters, Gaussian, Laplacian and edge detectors) to identify textures on the original image and at least one variant of the original. To generate the variants, entropy filters with different radius of influence and Gaussian blur are used; once this activity is completed, the automatic cropping technique is implemented, that is, the calculated textures are used, and the k-means algorithm is used with a k value of at least 2 to identify the pixels that belong to the background from those that belong to the embryo or any instrument or material present in the photograph that is not the embryo. Based on this 2, 3, or 4-dimensional mask, the edges are detected to crop the image containing the embryo. As an alternative to the automatic cropping based on the k-means algorithm, it is also possible to use an artificial intelligence-based object identifier, to identify and subtract from the image instruments, letters, or other artifacts unrelated to the embryo. To identify the stages of embryo development, a deep convolutional neural network model that can classify embryos into one of three stages is used: a) expanding, b) hatching and c) hatched; or it can function as a regressor defining the percentage that is inside the zona pellucida and the percentage that is outside. With the classifier technique, a probability value [0-1] is obtained, and the image corresponds to one of these three classes with an accuracy depending on the model used. To identify embryos that are collapsed (a natural process of embryos), a previously trained classifier can be used, which is identified with a probability index, or through a regressor that allows the identification of the percentage of collapse shown by the image of the embryo; to identify the degree of degradation of an embryo, one can use a pre-trained classifier that uses a probability index that a given image is in the “degraded’ or “normal” class, or a pre-trained regressor that identifies the percentage of degradation of the embryo through the image. To identify the degree of development of an embryo within a developmental curve, a previously trained classifier can be used, which uses a probability index indicating that a given image falls within one or several classes associated with the degree of evolution or growth of the embryo according to the expected growth given the embryo's conditions; as an alternative to the classifier, a previously trained regressor can be used to identify the percentile in which the embryo is located according to its growth and development, based on healthy embryos, or alternatively, statistical data can be used to locate the image of the embryo within a distribution of statistical parameters of growth such as the size of the different zones of the embryo. Therefore, at the end of this stage, a computer-implemented algorithm is available to identify the pixels or voxels that belong to at least one of the following five areas: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; the supervised training is performed through manual labels on the texture vectors for each pixel or voxel; where, from the predictions made at the pixel or voxel level, the predictions are subjected to a process to generate more homogeneous areas for each label, which involves the extraction of the blobs of a k-means (k=20) and a process of erosion and dilation of the zones. An alternative to identify the different zones is to use a neural network model containing an encoder and a decoder that associates each pixel or voxel to one of five labels: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis--, in [0060]-[0062]; also see SHAFIEE: e.g., -- the convolutional neural network 104 is trained on a plurality of images of embryos taken on a third day of embryo development…--, in [0023]-[0025]; and
{for b) a second model {a predictive model} that generate a second clinical parameter from a set of classification parameters that not derived from imaging of the embryo}; see SHAFIEE: e.g., -- another expert system (not shown). In practice, any of a variety of experts systems can be utilized in combination with the convolutional neural network, including support vector machines, random forest, self-organized maps, fuzzy logic systems, data fusion processes, ensemble methods, rule based systems, genetic algorithms, and artificial neural networks. It will be appreciated that the additional expert system may be trained on features from multiple stages of embryonic development as well as with features that are external to the images, such as biometric parameters of an egg donor, a sperm donor, or a recipient of the embryo.--, in [0026]; and, -- the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]); and
generating a composite parameter, representing a likelihood of aneuploidy in the embryo, from the first clinical parameter, the second clinical parameter, and the third clinical parameter (see Chavez Badiola: e.g., -- defined mathematical models that allow the identification of intensity patterns in 2, 3 or 4 dimensions such as roughness, contrast, brightness, saturation, smoothness, or particular shapes, for the prediction of embryo characteristics such as degree of collapse, or degradation, as well as the stage of the embryo; said parameters may be based on filters that identify a plurality of textures and/or other metrics based on segmentation of cell types: in which said plurality of textures may be at least one of those where a combination of texture detector masks are used such as 2-dimensional LAWS (Preferably 25, wherein the LAWS energy is obtained by detecting textures in embryo images which may be the standardized, or enhanced through machine vision strategies such as energy filters, Gaussian, Laplacian and edge detectors) to identify textures on the original image and at least one variant of the original. To generate the variants, entropy filters with different radius of influence and Gaussian blur are used; once this activity is completed, the automatic cropping technique is implemented, that is, the calculated textures are used, and the k-means algorithm is used with a k value of at least 2 to identify the pixels that belong to the background from those that belong to the embryo or any instrument or material present in the photograph that is not the embryo. Based on this 2, 3, or 4-dimensional mask, the edges are detected to crop the image containing the embryo. As an alternative to the automatic cropping based on the k-means algorithm, it is also possible to use an artificial intelligence-based object identifier, to identify and subtract from the image instruments, letters, or other artifacts unrelated to the embryo. To identify the stages of embryo development, a deep convolutional neural network model that can classify embryos into one of three stages is used: a) expanding, b) hatching and c) hatched; or it can function as a regressor defining the percentage that is inside the zona pellucida and the percentage that is outside. With the classifier technique, a probability value [0-1] is obtained, and the image corresponds to one of these three classes with an accuracy depending on the model used. To identify embryos that are collapsed (a natural process of embryos), a previously trained classifier can be used, which is identified with a probability index, or through a regressor that allows the identification of the percentage of collapse shown by the image of the embryo; to identify the degree of degradation of an embryo, one can use a pre-trained classifier that uses a probability index that a given image is in the “degraded’ or “normal” class, or a pre-trained regressor that identifies the percentage of degradation of the embryo through the image. To identify the degree of development of an embryo within a developmental curve, a previously trained classifier can be used, which uses a probability index indicating that a given image falls within one or several classes associated with the degree of evolution or growth of the embryo according to the expected growth given the embryo's conditions; as an alternative to the classifier, a previously trained regressor can be used to identify the percentile in which the embryo is located according to its growth and development, based on healthy embryos, or alternatively, statistical data can be used to locate the image of the embryo within a distribution of statistical parameters of growth such as the size of the different zones of the embryo. Therefore, at the end of this stage, a computer-implemented algorithm is available to identify the pixels or voxels that belong to at least one of the following five areas: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; the supervised training is performed through manual labels on the texture vectors for each pixel or voxel; where, from the predictions made at the pixel or voxel level, the predictions are subjected to a process to generate more homogeneous areas for each label, which involves the extraction of the blobs of a k-means (k=20) and a process of erosion and dilation of the zones. An alternative to identify the different zones is to use a neural network model containing an encoder and a decoder that associates each pixel or voxel to one of five labels: i) background, ii) zona pellucida, iii) trophectoderm, iv) blastocele and/or v) inner cell mass; [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]; also see SHAFIEE: e.g., -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; and, --the genetic algorithm 204 can also use biometric parameters representing of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg in determining a quality of the embryo. Further, the genetic algorithm 204 can utilize the outputs of multiple neural networks in determining the embryo quality, which can include other convolutional neural networks, for example, receiving as input an image of the embryo on a different day of development than the convolutional neural network 202, recurrent neural networks, capsule networks, and generative adversarial networks.--, in [0027]; and, --[0032] The system 300 includes an imager 306 that acquires an image of for each embryo on at least one selected day of development. This image is then provided to the convolutional neural network 302. It will be appreciated, however, that multiple images, taken from different days, can be used at one or multiple convolutional neural networks. The convolutional neural network 302 generates a plurality of values representing the morphology of the embryo from the provided one or more images. This plurality of values, as well as a plurality of features representing biometric parameters of one of a patient receiving the embryo, a sperm donor who provided sperm used to create the embryo, an egg utilized to produce the human embryo, and an egg donor who provided the egg to the expert system, are provided to the expert system 304. The biometric parameters can include, for example, an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. In the illustrated implementation, the biometric parameters are retrieved from an electronic health records (EHR) database 308.--, in [0032]).
Re Claim 17, Chavez Badiola as modified by SHAFIEE further disclose wherein generating the composite parameter from the first clinical parameter, the second clinical parameter, and the third clinical parameter comprises generating the composite parameter according to a majority vote among the first clinical parameter, the second clinical parameter, and the third clinical parameter (see SHAFIEE: e.g., -- [0029] A second training set of images, representing embryos of known quality, can be generated at the imager 206 and provided to the trained convolutional neural network 202 to generate a set of convolutional neural network outputs and known embryo qualities. For example, the known quality can be an outcome of the implantation or an average {e.g., arithmetic mean or median) of numerical grades provided via manual assessment by a set of embryologists from different institutions. This set of outputs and the embryo qualities can be provided as training data to the genetic algorithm.--, in [0029]; also see: -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]; also see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]; and, [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]).
Re Claim 18, Chavez Badiola as modified by SHAFIEE further disclose wherein the first predictive model is implemented as a support vector machine, and the second predictive model is implemented as a fully-connected feedforward neural network (see SHAFIEE: e.g., -- [0029] A second training set of images, representing embryos of known quality, can be generated at the imager 206 and provided to the trained convolutional neural network 202 to generate a set of convolutional neural network outputs and known embryo qualities. For example, the known quality can be an outcome of the implantation or an average {e.g., arithmetic mean or median) of numerical grades provided via manual assessment by a set of embryologists from different institutions. This set of outputs and the embryo qualities can be provided as training data to the genetic algorithm.--, in [0029]; also see: -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]; also see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]; and, [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]; and, -- provides the set of pre-processed images of the embryo, classified, and ranked to an expert system characterized in that the expert system is a feedforward neural network.--, in claim 6; also see: -- [0077] The proposed system then uses the previously described list of characteristics and feeds them into a previously trained AI model to predict the prognosis of each embryo. This resulted in a list of probability values that each list of characteristics associated with each embryo image belongs to the good prognosis class. The embryos are ranked in descending order according to their probability value of belonging to the good prognosis class using letters of the alphabet in order, therefore the letter ‘A’ is assigned to the embryo with the best prognosis.--, in [0077])
Re Claim 19, Chavez Badiola as modified by SHAFIEE further disclose wherein generating the second clinical parameter at the first predictive model from the first subset of the set of parameters comprises generating the second clinical parameter at the first predictive model from the set of parameters, and generating the third clinical parameter at the second predictive model from the second subset of the set of parameters comprises generating the third clinical parameter at the second predictive model from the set of parameters (see SHAFIEE: e.g., -- [0029] A second training set of images, representing embryos of known quality, can be generated at the imager 206 and provided to the trained convolutional neural network 202 to generate a set of convolutional neural network outputs and known embryo qualities. For example, the known quality can be an outcome of the implantation or an average {e.g., arithmetic mean or median) of numerical grades provided via manual assessment by a set of embryologists from different institutions. This set of outputs and the embryo qualities can be provided as training data to the genetic algorithm.--, in [0029]; also see: -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]; also see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]; and, [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]).
Re Claim 20, Chavez Badiola as modified by SHAFIEE further disclose wherein each of the first subset of the set of parameters and the second subset of the set of parameters are proper subsets of the set of parameters (see SHAFIEE: e.g., -- [0029] A second training set of images, representing embryos of known quality, can be generated at the imager 206 and provided to the trained convolutional neural network 202 to generate a set of convolutional neural network outputs and known embryo qualities. For example, the known quality can be an outcome of the implantation or an average {e.g., arithmetic mean or median) of numerical grades provided via manual assessment by a set of embryologists from different institutions. This set of outputs and the embryo qualities can be provided as training data to the genetic algorithm.--, in [0029]; also see: -- in addition to the morphological features provided by the neural network, a plurality of features representing biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg can be provided to the expert system for use in determining the output class. These biometric parameters can include at least one of an age of the egg, a body mass index of the patient, an age of the patient, an egg maturation status, a method of fertilization for the embryo, a treatment regime for the patient, a hormonal profile of the patient, and an age of the egg donor, a past diagnosis of a condition of the patient, a past diagnosis of a condition of the sperm donor, and an endometrium thickness of the patient. For example, the expert system can be implemented as a feedforward neural network. In another implementation, the expert system includes a genetic algorithm that calculates a plurality of weights corresponding to the plurality of values, such that the output is determined as weighted linear combination of the plurality of values.--, in [0037]; also see: -- [0030] The output of the genetic algorithm 204 is a set of weights that optimally predict the quality of the embryo from the outputs of the convolutional neural network 202. This output can be provided to a user at an associated user interface 208. The inventors have found that a combination of the convolutional neural network 202 with a genetic algorithm 204 trained in this manner can provide predictions of implantation outcomes that meet or exceed that of experienced embryologists. As a result, the selection of embryos for transfer can be automated without any significant loss of accuracy.--, in [0029]-[0030], and, -- [0034] In one implementation, the expert system 304 is implemented as a feed-forward neural network. In this approach, the convolutional neural network 302 is pre-trained on training images and the weights within the network are frozen. The feed-forward neural network receives the biometric parameters at an input layer, with the output of the convolutional neural network is injected into a hidden layer of the feedforward neural network. A final layer of the feed-forward neural network can be implemented as a softmax layer to provide a classification result. The feed-forward neural network is trained on embryo images and a known quality for each training image to provide the illustrated system.--, in [0034]-[0035]; also see Chavez Badiola: e.g., --The genetic results showed that embryo 2 was the only aneuploid, indicating that, if the embryo had been selected using the criteria of the embryology team, the aneuploid would have been selected, and the procedure would have been unsuccessful. On the other hand, the system identified embryo 2 as the embryo with the lowest probability of having a good prognosis.--, in [0080], and, --a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or some other algorithm can be used. Understanding good prognosis as an egg that successfully achieved fertilization, or that developed into a euploid embryo and/or transferred with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis such as an egg that did not mature at the meiosis I to meiosis II stage and did not achieve normal fertilization, or one that generated an aneuploid embryo with b-hCG value <20 and/or miscarriage after the embryo implantation.--, in [0094]; and, [0061] 3. ASSIGNING FEASIBILITY POTENTIAL. With the algorithm of the previous stage, in addition to the descriptors related to the collapse and degradation phases, other descriptors based on the following are obtained: original image, image with entropy filters, image with highlighted edges, polar image from a centroid, and the areas identified by the segmentation methodology. Statistical descriptors are used, or a fraction of them selected by descriptor selection methods, associated with the distribution of the data, including but not limited to measures of central tendency, dispersion, and kurtosis; so that with the list of descriptors obtained from each egg, together with the history of the patient (age and hours between fertilization and the image or images), and the source of the egg (laboratory preset: microscope, and objective), a deep neural network is trained to classify each embryo into one of two classes: good prognosis and poor prognosis. Alternatively, another classification algorithm such as support vector machines, decision trees, or other algorithm can be used. Understanding good prognosis as a euploid and/or transferred embryo with (beta-human chorionic gonadotropin) b-hCG>=20 units (beta positive, pregnancy, 7 days after transfer) and/or presence of gestational sac at least three weeks after transfer, observed by imaging techniques and/or presence of heartbeat at least five weeks after transfer and/or evidence of live birth; and one with a poor prognosis as an aneuploid or b-hCG value <20 and/or miscarriage after embryo implantation. If there is a conflict in the criteria used to classify embryo images (for example euploid with b-hCG<20), priority is given to the ploidy level of the embryo;
[0062] Finally, a set of embryos is ranked in descending (or ascending) order according to the probability of having a good prognosis; in such a way that the health care team evaluates the results obtained by the algorithm, together with the patient's history and decides which embryos will be transferred, depending on the case.--, in [0060]-[0062]).
Conclusion
Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to WEI WEN YANG whose telephone number is (571)270-5670. The examiner can normally be reached on 8:00 - 5:00 pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Amandeep Saini can be reached on 571-272-3382. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/WEI WEN YANG/Primary Examiner, Art Unit 2662
Prosecution Insights