DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. IB2023/055761, filed on June 5, 2023.
Response to Amendment
In the amendment filed on May 7, 2026, the following has occurred: claim(s) 1, 4, 6, 9, 12-13, 15, 17, 20 have been amended. Now, claim(s) 1-20 are pending.
Notice to Applicant
The Examiner has withdrawn the 35 U.S.C. 101 rejections on the claims 1-20 as the claimed limitations recite additional element(s) demonstrating that the claims as a whole integrates the judicial exception into a practical application. The Applicant’s newly amended claimed limitation of “output the image data representing the one or more predicted images of the future appearance of the wound, wherein the image data for each of the one or more predicted images comprises pixel data representative of the future appearance of the wound at the corresponding future time” by “pass the image capture data for the sequence of the one or more images through a machine learning model trained to generate image data representing one or more predicted images of a future appearance of the wound, each of the one or more predicted images representative of the future appearance of the wound at a corresponding future time, the machine learning model trained using historical image data” , and one way to demonstrate that a claim is directed to patent-eligible subject matter is by applying or using the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception (See MPEP 2106.05(e)). The claims 1-20 recite additional elements that integrate the recited judicial exception into a practical application. The 35 U.S.C. 101 rejection(s) have been withdrawn.
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-3, 9-14 are rejected under 35 U.S.C. 102(a)(1) and (a)(2) as being unpatentable over Fan et al. (U.S. Patent Pre-Grant Publication No. 2021/0201479).
As per independent claim 1, Fan discloses a system comprising:
a memory (See [0173]: Any of the machine learning systems and/or methods of the present technology may be implemented on or in communication with processors and/or memory of the various imaging systems and devices of the present disclosure); and
a processing unit including one or more processors coupled to the memory, the one or more processors (See [0173]: Any of the machine learning systems and/or methods of the present technology may be implemented on or in communication with processors and/or memory of the various imaging systems and devices of the present disclosure) configured to execute instructions that cause the processing unit to:
obtain image capture data for a sequence of one or more images (See [0116]-[0117]: The processor along with capture control module 1135, multi-aperture spectral camera, and working memory 1105 represent one means for capturing a set of spectral images and/or a sequence of images) representative of an appearance of a wound at (See [0138]: Where PPG information is included, the disclosed imaging systems provide a method to assess pathologies involving changes to tissue blood flow and pulse rate including: tissue perfusion; cardiovascular health; wounds such as ulcers; peripheral arterial disease, and respiratory health) a corresponding image capture time, each of the one or more images prior to a final image of the sequence of one or more images separated by a sampling time interval between the image and a next image (See [0138]-[0139]: A number of different images at the same wavelength corresponding to different times (PPG data), which the Examiner is interpreting to encompass a corresponding image capture time, each of the one or more images prior to a final image of the sequence of one or more images separated by a sampling time interval between the image and a next image as the last image captured would be the final image),
pass the image capture data for the sequence of one or more images through a machine learning model trained to generate image data representing one or more predicted images of a future appearance of the wound (See [0140]-[0144]: The multispectral datacube can be analyzed as input data into a machine learning model to generate a classified mapping of the imaged tissue, which the Examiner is interpreting a machine learning model to encompass a machine learning model trained to generate image data representing one or more predicted images of a future appearance of the wound as the machine learning model can be an artificial neural network ([0141]), and the training data can include multispectral datacubes (the input) and classified mappings (the expected output) that have been labeled, for example by a clinician who has designated areas of the wound that correspond to certain clinical states ([0144])), each of the one or more predicted images representative of the future appearance of the wound at a corresponding future time (See [0140]-[0144]: Other implementations of the machine learning model can be trained to make other types of predictions, for example the likelihood of a wound healing to a particular percentage area reduction over a specified time period (e.g., at least 50% area reduction within 30 days) or wound states such as, hemostasis, inflammation, pathogen colonization, proliferation, remodeling or healthy skin categories, which the Examiner is interpreting other types of predictions to encompass each of the one or more predicted images representative of the future appearance of the wound at a corresponding future time), the machine learning model trained using historical image data, the historical image data comprising one or more historical image data sets, each historical image data set of the one or more historical image data sets comprising image data for a historical sequence of images of an appearance of a corresponding historical wound (See Table 1, [0172]: The disclosed machine learning systems can learn to determine wound healing potential through being exposed to large volumes of labeled training data, which the Examiner is interpreting the labeled training data to encompass the machine learning model trained using historical image data), wherein a prediction time interval between the corresponding future time and a capture time of a last image of the one or more sequence of the one or more images is greater than each of the sampling time intervals (See [0138]-[0139]: A number of different images at the same wavelength corresponding to different times (PPG data), which the Examiner is interpreting the final time of the PPG data to encompass a capture time of a last image of the one or more sequence of one or more images is greater than each of the sampling time intervals as the final time of the PPG data would be greater than the earlier times), and
output the image data representing the one or more predicted images of the future appearance of the wound (See Fig. 24-25, [0161]-[0162]: This can in turn be provided to a machine learning classifier, for example a fully connected feedforward artificial neural network or the system shown in FIG. 25, in order to output a healing prediction for the imaged ulcer or other wound, which the Examiner is interpreting output a healing prediction for the imaged ulcer or other wound to encompass the claimed portion as a healing prediction of an ulcer is an example of a future appearance of a wound), wherein the image data for each of the one or more predicted images comprises pixel data representative of the future appearance of the wound at the corresponding future time (See Fig. 33, [0008]-[0011], [0196]-[0197]: To accomplish this output, a machine learning algorithm was trained to take MSI or RGB data as input and generate predicted healing parameters for portions of the wound (e.g., for individual pixels or subsets of pixels in a wound image), which the Examiner is interpreting predicted healing parameters for portions of the wound (e.g., for individual pixels or subsets of pixels in a wound image) to encompass pixel data representative of the future appearance of the wound at the corresponding future time as an image of the wound to be displayed to the user such that the healing pixels and the non-healing pixels are displayed in different visual representations.)
Claim 13 mirrors claim 1 only within a different statutory category, and is rejected for the same reason as claim 1.
As per claim 2, Fan discloses the system of claim 1 as described above. Fan further teaches wherein the image capture data includes metadata identifying a treatment method or one or more treatment method parameters (See [0200]-[0201]: The results of a layer within the convolutional neural network can be modified by information from another source, clinical data from a subject's medical history or treatment plan (e.g., patient health metrics or clinical variables as described herein) can be used as the source of this modification, which the Examiner is interpreting treatment plan to encompass metadata identifying a treatment method or one or more treatment method parameters.)
As per claim 3, Fan discloses the system of claims 1-2 as described above. Fan further teaches wherein the treatment method parameters include negative-pressure wound therapy (NPWT) parameters (See [0157]-[0159]: A wound assessment system or a clinician can determine an appropriate level of wound care therapy based on the results of the machine learning algorithms, the AWC therapies include negative-pressure wound therapy.)
As per claim 9, Fan discloses the system of claim 1 as described above. Fan further teaches wherein:
the machine learning model comprises a second machine learning model (See [0191]-[0192]: The compressed image vector was used as an input to a second supervised machine learning algorithm);
a first machine learning model is trained prior to the second machine learning model using a first training image data set (See [0189]-[0192]: Upon extraction of the compressed image vector, the compressed image vector was used as an input to a second supervised machine learning algorithm, which the Examiner is interpreting the input to a second machine learning algorithm to encompass a first machine learning model is trained prior to the second machine learning model using a first training image data set as the identical encoder-decoder algorithm (interpreting to encompass first trained algorithm) was used for all images in the data set) that includes a first subset of images of the historical sequence of images captured during a sampling period associated with the corresponding historical wound and a second subset of images captured during the sampling period, wherein a number of images in the first subset of images is greater than the number of images in the second subset of images (See [0168]-[0170]: The dimensionality reduction allows the autoencoder neural network to learn the most salient features of the input images, where the innermost layer (or another inner layer) of the autoencoder represents a “feature reduction” version of the input, this can serve to reduce an image having, for example, approximately 1 million pixels (where each pixel value can be considered as a separate feature of the image) to a feature set of around 50 values, and the reduced-dimensionality representation of the images can be used by another machine learning model, which the Examiner is interpreting the reduced images to encompass a number of images in the first subset of images is greater than the number of images in the second subset of images); and
the second machine learning model is constrained to use one or more layers of the first machine learning model (See [0168]-[0169]: A fully connected neural network is one in which each node in the input layer is connected to each node in the subsequent layer (the first hidden layer), each node in that first hidden layer is connected in turn to each node in the subsequent hidden layer, and so on until each node in the final hidden layer is connected to each node in the output layer.)
As per claim 10, Fan discloses the system of claims 1 and 9 as described above. Fan further teaches wherein the first machine learning model is trained bi-directionally (See [0167]-[0169]: An artificial neural network may be an adaptive system that is configured to change its structure (e.g., the connection configuration and/or weights) based on information that flows through the network during training, and the weights of the hidden layers can be considered as an encoding of meaningful patterns in the data, which the Examiner is interpreting the encoder/decoder to encompass bi-directionally.)
As per claim 11, Fan discloses the system of claim 1 as described above. Fan further teaches wherein the prediction time interval is greater than an input time interval associated with the sequence of images (See [0011]-[0012]: Determining the predicted healing parameter over the predetermined time interval, which the Examiner is interpreting the predetermined time interval to encompass the prediction time interval is greater than an input time interval as the predetermined time interval can be 30 days (Claim 4).)
As per claim 12, Fan discloses the system of claim 1 as described above. Fan further teaches wherein the machine learning model is trained using historical metadata corresponding to the historical image sequence (See [0070], [0144]: During training, an artificial neural network can be exposed to pairs in its training data and can modify its parameters to be able to predict the output of a pair when provided with the input, the training data can include multispectral datacubes (the input) and classified mappings (the expected output) that have been labeled, for example by a clinician who has designated areas of the wound that correspond to certain clinical states, and/or with healing (1) or non-healing (0) labels sometime after initial imaging of the wound when actual healing is known), and wherein the processing unit is further configured to:
obtain metadata comprising wound properties of the wound corresponding to the sequence of images, the wound properties comprising one or more of wound area, wound depth, or wound healing stage (See [0070], [0144], [0165]-[0166]: During training, an artificial neural network can be exposed to pairs in its training data and can modify its parameters to be able to predict the output of a pair when provided with the input, the training data can include multispectral datacubes (the input) and classified mappings (the expected output) that have been labeled, for example by a clinician who has designated areas of the wound that correspond to certain clinical states, and/or with healing (1) or non-healing (0) labels sometime after initial imaging of the wound when actual healing is known, which the Examiner is interpreting areas of the wound that correspond to certain clinical states to encompass the wound properties comprising one or more of wound area, wound depth, or wound healing stage);
pass the metadata through the machine learning model to generate predicted metadata for the wound at the corresponding future time (See [0141]-[0144]: These metrics can be converted into a vector representation through appropriate processing, for example through word-to-vec embeddings, a vector having binary values representing whether the patient does or does not have the patient metric (e.g., does or does not have type I diabetes), or numerical values representing a degree to which the patient has each patient metric, which the Examiner is interpreting the metrics can be converted into a vector to encompass pass the metadata through the machine learning model as the final hidden layer is connected to each node in the output layer), the predicted metadata comprising one or more predicted wound properties of the wound at the corresponding future time (See [0160]-[0164]: Wound assessment and/or healing predictions described herein may be accomplished based on one or more images of the wound, either alone or based on a combination of both patient health data (e.g., one or more health metric values, clinical features, etc.) and images of the wound, and patient metrics or a combination of some or all of the patient metrics to improve the accuracy of predicted healing parameters generated by the systems and methods, which the Examiner is interpreting predicted healing parameters to encompass one or more predicted wound properties of the wound at the corresponding future time); and
output the predicted metadata (See [0141]-[0144]: The classified mappings are the expected output, which the Examiner is interpreting to encompass the claimed portion.)
As per claim 14, Fan discloses the method of claim 13 as described above. Fan further teaches wherein the machine learning model is trained using a weighted loss that assigns a first weight to a first image that is less than a second weight assigned to a second image having a corresponding predicted future time that is later than the predicted future time corresponding to the first image (See [0141]-[0143], [0169]-[0170]: The nodes in each convolutional layer of a CNN can share weights such that the convolutional filter of a given layer is replicated across the entire width and height of the input volume (e.g., across an entire frame), reducing the overall number of trainable weights and increasing applicability of the CNN to data sets outside of the training data, and the values of a layer may be pooled to reduce the number of computations in a subsequent layer (e.g., values representing certain pixels may be passed forward while others are discarded), and further along the depth of the CNN pool masks may reintroduce any discarded values to return the number of data points to the previous size, which the Examiner is interpreting the weights to encompass first weight to a first image that is less than a second weight assigned to a second image, and interpreting the algorithm can take data from an RGB image, and optionally the subject's medical history or other clinical variables, and output a predicted healing parameter such as a conditional probability that indicates whether the DFU will respond to 30 days of standard wound care therapy ([0184]) to encompass a second image having a corresponding predicted future time that is later than the predicted future time corresponding to the first image as a second image could predict a further future time than 30 days if the second image was taken at a later time, and the Examiner is interpreting the values of a layer may be pooled to reduce the number of computations in a subsequent layer to encompass the weighted loss.)
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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, 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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 4-8, 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Fan et al. (U.S. Patent Pre-Grant Publication No. 2021/0201479) in view of Park et al. (U.S. Patent Pre-Grant Publication No. 2022/0101984).
As per claim 4, Fan discloses the system of claim 1 as described above. Fan further teaches wherein the machine learning model is trained bi-directionally, wherein a first direction of training trains the machine learning model to generate the one or more predicted images from the historical sequence of images and wherein a second direction of training trains the machine learning model to generate a reconstructed first image from the one or more predicted images and images in the historical sequence of images subsequent to the first reconstructed image (See [0167]-[0169]: An artificial neural network may be an adaptive system that is configured to change its structure (e.g., the connection configuration and/or weights) based on information that flows through the network during training, and the weights of the hidden layers can be considered as an encoding of meaningful patterns in the data, which the Examiner is interpreting the encoder to encompass a first direction of training trains the machine learning model to generate the one or more predicted future images from the historical sequence of images, and interpreting the decoder to encompass a second direction of training trains the machine learning model to generate a reconstructed first image from the one or more predicted images and images in the historical sequence of images subsequent to the first reconstructed image as the goal of certain autoencoders is to compress the input data with the encoder, then decompress this encoded data with the decoder such that the output is a good/perfect reconstruction of the original input data), wherein the first direction of training and the second direction of training are used to adjust weights in layers of the machine learning model.
While Fan discloses a system wherein the machine learning model is trained bi-directionally, wherein a first direction of training trains the machine learning model to generate the one or more predicted images from the historical sequence of images and wherein a second direction of training trains the machine learning model to generate a reconstructed first image from the one or more predicted images and images in the historical sequence of images subsequent to the first reconstructed image, Fan may not explicitly teach wherein the first direction of training and the second direction of training are used to adjust weights in layers of the machine learning model.
Park teaches a system wherein the first direction of training and the second direction of training are used to adjust weights in layers of the machine learning model (See [0068], [0140]: The model learning unit 114 can obtain an error between the output label information and the ground truth label information, and update a weight of the machine learning model while propagating the error backwards, which the Examiner is interpreting update a weight of the machine learning model to encompass adjust weights in layers of the machine learning model as the algorithms that can be used can include bidirectional recurrent deep neural network (BRDNN).)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed to modify the system of Fan to include the first direction of training and the second direction of training are used to adjust weights in layers of the machine learning model as taught by Park. One of ordinary skill in the art before the effective filing date of the claimed invention would have been motivated to modify Fan with Park with the motivation of normalizing data for machine learning (See Background Art of Park in Paragraph [0004]).
Claim 15 mirrors claim 4 only within a different statutory category, and is rejected for the same reason as claim 4.
As per claim 5, Fan/Park discloses the system of claims 1 and 4 as described above. Fan further teaches wherein layers in the machine learning model are shared by the first direction of training and the second direction of training (See [0168]-[0169]: A fully connected neural network is one in which each node in the input layer is connected to each node in the subsequent layer (the first hidden layer), each node in that first hidden layer is connected in turn to each node in the subsequent hidden layer, and so on until each node in the final hidden layer is connected to each node in the output layer.)
Claim 16 mirrors claim 5 only within a different statutory category, and is rejected for the same reason as claim 5.
As per claim 6, Fan discloses the system of claim 1 and Fan/Park discloses the system of claim 4 as described above. Fan further teaches wherein:
the machine learning model comprises a second machine learning model (See [0191]-[0192]: The compressed image vector was used as an input to a second supervised machine learning algorithm);
a first machine learning model is trained prior to the second machine learning model using a first training image data set (See [0189]-[0192]: Upon extraction of the compressed image vector, the compressed image vector was used as an input to a second supervised machine learning algorithm, which the Examiner is interpreting the input to a second machine learning algorithm to encompass a first machine learning model is trained prior to the second machine learning model using a first training image data set as the identical encoder-decoder algorithm (interpreting to encompass first trained algorithm) was used for all images in the data set) that includes a first subset of images of the historical sequence of images captured during a sampling period associated with historical wound images and a second subset of images captured after the sampling period (See [0190]-[0192]: The identical encoder-decoder algorithm was used for all images in the data set, which the Examiner is interpreting all images in the data set to encompass a first training image data set that includes a first subset of images of the historical sequence of images captured during a sampling period associated with historical wound images and a second subset of images captured after the sampling period), wherein the second subset of images includes one or more images of the corresponding historical wound captured during a treatment period after the sampling period (See [0012], [0144], [0160], [0170]: To accomplish the prediction, a machine learning algorithm was trained to take MSI data and clinical variables as inputs and to output a scalar value representing the predicted PAR, which the Examiner is interpreting a machine learning algorithm was trained to take MSI data and clinical variables as inputs to encompass the second subset of images includes one or more images of the corresponding historical wound captured during a treatment period after the sampling period as during training, an artificial neural network can be exposed to pairs in its training data and can modify its parameters to be able to predict the output of a pair when provided with the input, which the Examiner is interpreting the pairs in the training data to encompass the corresponding historical wound captured during a treatment period after the sampling period); and
the second machine learning model is constrained to include one or more layers of the first machine learning model (See [0168]-[0169]: A fully connected neural network is one in which each node in the input layer is connected to each node in the subsequent layer (the first hidden layer), each node in that first hidden layer is connected in turn to each node in the subsequent hidden layer, and so on until each node in the final hidden layer is connected to each node in the output layer.)
Claim 17 mirrors claim 6 only within a different statutory category, and is rejected for the same reason as claim 6.
As per claim 7, Fan discloses the system of claim 1 and Fan/Park discloses the system of claims 4 and 6 as described above. Fan further teaches wherein the first machine learning model is trained bi-directionally (See [0167]-[0169]: An artificial neural network may be an adaptive system that is configured to change its structure (e.g., the connection configuration and/or weights) based on information that flows through the network during training, and the weights of the hidden layers can be considered as an encoding of meaningful patterns in the data, which the Examiner is interpreting the encoder/decoder to encompass bi-directionally.)
Claim 18 mirrors claim 7 only within a different statutory category, and is rejected for the same reason as claim 7.
As per claim 8, Fan discloses the system of claim 1 and Fan/Park discloses the system of claims 4 and 6 as described above. Fan further teaches wherein the one or more layers comprise a final layer, penultimate layer, or one or more mid-level layers (See [0168]: A fully connected neural network is one in which each node in the input layer is connected to each node in the subsequent layer (the first hidden layer), each node in that first hidden layer is connected in turn to each node in the subsequent hidden layer, and so on until each node in the final hidden layer is connected to each node in the output layer.)
As per claim 19, Fan discloses the method of claim 13 and Fan/Park discloses the method of claims 15 and 17 as described above. Fan further teaches wherein the one or more layers comprise a final layer (See [0168]: A fully connected neural network is one in which each node in the input layer is connected to each node in the subsequent layer (the first hidden layer), each node in that first hidden layer is connected in turn to each node in the subsequent hidden layer, and so on until each node in the final hidden layer is connected to each node in the output layer.)
As per claim 20, Fan discloses the method of claim 13 and Fan/Park discloses the method of claim 15 as described above. Fan further teaches wherein:
the machine learning model comprises a second machine learning model (See [0191]-[0192]: The compressed image vector was used as an input to a second supervised machine learning algorithm);
a first machine learning model is trained prior to the second machine learning model using a first training image data set (See [0189]-[0192]: Upon extraction of the compressed image vector, the compressed image vector was used as an input to a second supervised machine learning algorithm, which the Examiner is interpreting the input to a second machine learning algorithm to encompass a first machine learning model is trained prior to the second machine learning model using a first training image data set as the identical encoder-decoder algorithm (interpreting to encompass first trained algorithm) was used for all images in the data set) that includes a first subset of images of the historical sequence of images captured during a sampling period associated with the historical wound and a second subset of images captured during a treatment period of the wound, wherein a number of images in the first subset of images is greater than the number of images in the second subset of images (See [0168]-[0170]: The dimensionality reduction allows the autoencoder neural network to learn the most salient features of the input images, where the innermost layer (or another inner layer) of the autoencoder represents a “feature reduction” version of the input, this can serve to reduce an image having, for example, approximately 1 million pixels (where each pixel value can be considered as a separate feature of the image) to a feature set of around 50 values, and the reduced-dimensionality representation of the images can be used by another machine learning model, which the Examiner is interpreting the reduced images to encompass a number of images in the first subset of images is greater than the number of images in the second subset of images); and
the second machine learning model is constrained to use one or more layers of the first machine learning model (See [0168]-[0169]: A fully connected neural network is one in which each node in the input layer is connected to each node in the subsequent layer (the first hidden layer), each node in that first hidden layer is connected in turn to each node in the subsequent hidden layer, and so on until each node in the final hidden layer is connected to each node in the output layer.)
Response to Arguments
In the Remarks filed on May 7, 2026, the Applicant argues that the newly amended and/or added claims overcome the Claim Objection(s), Claim Interpretation(s), 35 U.S.C. 101 rejection(s), and 35 U.S.C. 102 rejection(s). The Examiner acknowledges that the newly added and/or amended claims overcome the Claim Objection(s), Claim Interpretation(s), 35 U.S.C. 101 rejection(s). However, the Examiner does not acknowledge that the newly added and/or amended claims overcome the 35 U.S.C. 102 rejection(s) and 35 U.S.C. 103 rejection(s).
The Applicant argues that:
(1) Fan does not disclose, teach, or suggest the claimed features. Fan does not generate image data representing predicted images of the future appearance of a wound. The Office Action contends that Fan's machine learning model generates ''other types of predictions, for example the likelihood of a wound healing to a particular percentage area reduction over a specified time period'' and that this encompasses ''each of the one or more predicted images representative of the future appearance of the wound at a corresponding future time." Office Action at pp. 14-15. The Office Action further contends that Fan's ''output a healing prediction for the imaged ulcer or other wound'' encompasses ''the image data representing the one or more predicted images of the future appearance of the wound." Office Action at p. 16. Applicant respectfully disagrees. Fan's machine learning models output classifications, scalar values, and probabilities about a current wound image. Specifically, Fan discloses a ''classified mapping'' that assigns each pixel in current image data to a tissue classification or healing potential score (Fan Paragraph [0140]), a ''scalar value'' representing a predicted healing parameter such as percent area reduction (Fan Paragraphs [0161], [0174]), a ''conditional probability'' indicating whether a wound will respond to therapy (Fan Paragraphs [0184]), and a binary ''healing'' or ''non-healing'' label (Fan Paragraph [0144]). Each of these is a classification, score, probability, or label derived from or applied to a current wound image. None is a new image depicting what the wound will look like at a future point in time. Amended claim 1 makes this distinction explicit. Claim 1 now recites that ''the image data for each of the one or more predicted images comprises pixel data representative of the future appearance of the wound at the corresponding future time." The output of the claimed system is pixel data, that is, image data in the form of an image depicting the predicted future appearance of the wound. This is fundamentally different from Fan's scalar values, probabilities, classifications, and mappings applied to current wound images;
(2) Fan does not disclose the claimed temporal structure. The Office Action maps the claimed sampling time intervals to Fan's PPG data, which involves ''a number of different images at the same wavelength corresponding to different times." Office Action at p. 14 (citing Fan Paragraphs [0138]-[0139]). The Office Action then asserts that ''the final time of the PPG data would be greater than the earlier times'' and that this encompasses the limitation that the prediction time interval is greater than each of the sampling time intervals. Office Action at p. 15. This mapping is incorrect. Fan's PPG data captures rapid physiological changes (pulsatile blood flow) over very short intervals. Fan Paragraph [0128]. The claimed wound image sequence captures wound appearance over days or weeks, with sampling time intervals between images. Spec. Paragraphs [0022], [0035]-[0039]. More fundamentally, the Examiner confuses the capture time of the last PPG image with the prediction time interval. The prediction time interval is the interval between the capture time of the last image in the sequence and the corresponding future time of the predicted image. It is not the capture time of the last image itself. Fan does not disclose a prediction time interval between a future time associated with a predicted image and the capture time of a last image, because Fan does not generate predicted images at future times. For at least these reasons, Fan does not disclose, teach, or suggest the features of amended independent claim 1. Reconsideration and withdrawal of the rejection of independent claim 1 are respectfully requested;
(3) independent claim 13 recites a method comprising, in part, the same features discussed above with respect to independent claim 1, including passing image capture data through a machine learning model trained to generate image data representing predicted images of a future appearance of the wound, outputting image data comprising pixel data representative of the future appearance of the wound at the corresponding future time, and the prediction time interval being greater than each of the sampling time intervals. For at least the same reasons discussed above, Fan does not disclose, teach, or suggest the features of amended independent claim 13. Accordingly, Applicant respectfully submits that independent claim 13 is in condition for allowance;
(4) claim 4 recites two distinct training directions with two distinct objectives. The first direction trains the model to generate predicted future images from a historical sequence. The second direction trains the model to reconstruct a past image from the predicted future images and subsequent historical images. Both directions are used to adjust weights in the model's layers. This bi-directional training structure, in which future prediction and past reconstruction each contribute to weight adjustment, is not disclosed by Fan's autoencoder, which compresses and reconstructs the same image without generating predicted future images in either direction. Accordingly, Fan does not disclose, teach, or suggest the features of amended dependent claim 4. Reconsideration and withdrawal of the rejection of claim 4, and claims 5, 7, and 8 depending therefrom, are respectfully requested;
(5) Fan does not disclose training the autoencoder using a data set that includes a first subset of images captured during a sampling period and a second subset of images captured during a treatment period after the sampling period, as claim 6 requires. Fan's autoencoder processes all images identically for dimensionality reduction without regard to whether images were captured during a sampling period or a subsequent treatment period. Furthermore, Fan's second supervised algorithm receives the compressed image vector as input. Fan Paragraph [0192]. Fan does not disclose that the second algorithm is ''constrained to include one or more layers'' of the autoencoder. Receiving a compressed vector as input is not the same as being constrained to include layers of a prior model. Accordingly, Fan does not disclose, teach, or suggest the features of amended dependent claim 6. Reconsideration and withdrawal of the rejection of claim 6 is respectfully requested;
(6) Fan's dimensionality reduction reduces the number of features (e.g., pixel values) representing a single image, not the number of images in a training subset. Fan Paragraph [0169]. Claim 9 requires that the first subset of images contains a greater number of images than the second subset of images. This is a requirement about the relative sizes of two subsets of training images, not about the dimensionality of feature representations. Fan does not disclose training a first model with two subsets of images where one subset contains more images than the other. Accordingly, Fan does not disclose, teach, or suggest the features of amended dependent claim 9. Reconsideration and withdrawal of the rejection of claim 9, and claim 10 depending therefrom, are respectfully requested;
(7) Fan uses clinical variables as input to a machine learning model to improve classification accuracy. Fan Paragraphs [0144], [0163]-[0164]. Fan does not disclose generating predicted metadata comprising predicted wound properties at a corresponding future time. Fan's model outputs healing predictions (scalar values, probabilities, classifications) about the current wound, not predicted future wound properties such as predicted wound area, wound depth, or wound healing stage at a future time. Amended claim 12 clarifies this distinction explicit by reciting that the predicted metadata comprises ''one or more predicted wound properties of the wound at the corresponding future time." Accordingly, Fan does not disclose, teach, or suggest the features of amended dependent claim 12. Reconsideration and withdrawal of the rejection of claim 12 is respectfully requested;
(8) dependent claims 15, 17, and 20 are method counterparts of claims 4, 6, and 9, respectively. For at least the same reasons discussed above with respect to claims 4, 6, and 9, Fan does not disclose, teach, or suggest the features of amended dependent claims 15, 17, and 20. Reconsideration and withdrawal of the rejections of claims 15, 17, and 20, and claim 18 depending from claim 17, are respectfully requested. The remaining dependent claims 2, 3, 5, 7, 8, 10, 11, 14, 16, 18, and 19 each depend from independent claim 1 or independent claim 13, discussed above, and are believed to be allowable for at least similar reasons. Because each dependent claim is deemed to define an additional aspect of the systems and methods provided herein, the individual consideration of each on its own merits is respectfully requested. Accordingly, reconsideration and withdrawal of the rejections of the dependent claims are respectfully requested.
In response to argument (1), the Examiner does not find the Applicant’s argument(s) persuasive. The Examiner maintains that Fan teaches the “generate image data representing predicted images of the future appearance of a wound” as described in Paragraph [0140]: “At block 1530, the multispectral datacube 1525 can be analyzed as input data 1525 into a machine learning model 1532 to generate a classified mapping 1535 of the imaged tissue. The classified mapping can assign each pixel in the image data (which, after registration, represent specific points on the imaged object 1511) to a certain tissue classification, or to a certain healing potential score. The different classifications and scores can be represented using visually distinct colors or patterns in the output classified image. Thus, even though a number of images are captured of the object 1511, the output can be a single image of the object (e.g., a typical RGB image) overlaid with visual representations of pixel-wise classification.” The Examiner maintains that under the broadest reasonable interpretation of “the image data for each of the one or more predicted images comprises pixel data representative of the future appearance of the wound at the corresponding future time” as data based on the prediction of the wound, which the Examiner is interpreting to be encompassed by Fan specifically in Paragraphs [0184]-[0186]. The 35 U.S.C. 102 rejection(s) stand.
In response to argument (2), the Examiner does not find the Applicant’s argument(s) persuasive. The Examiner maintains that Fan teaches “wound image sequence captures wound appearance over days or weeks, with sampling time intervals between images” as in Paragraph [0144] describes “Other implementations of the machine learning model 1532 can be trained to make other types of predictions, for example the likelihood of a wound healing to a particular percentage area reduction over a specified time period (e.g., at least 50% area reduction within 30 days) or wound states such as, hemostasis, inflammation, pathogen colonization, proliferation, remodeling or healthy skin categories.” and Fig. 24 and Paragraphs [0161]-[0163] describes “As illustrated, an image of a wound, or a set of multispectral images of the wound captured at different wavelengths, either at different times or simultaneously using a multispectral image sensor, may be used to provide both the input and output values to a neural network such as an autoencoder neural network, which is a type of artificial neural network as described in greater detail below.” The Examiner maintains that Fan teaches the Applicant’s claims as amended. The 35 U.S.C. 102 rejection(s) stand.
In response to argument (3), the Examiner does not find the Applicant’s argument(s) persuasive. The Examiner maintains that Fan encompasses newly amended independent claim 13 as similarly responded above in response to arguments (1) and (2). The 35 U.S.C. 102 rejection(s) stand.
In response to argument (4), the Examiner finds the Applicant’s argument(s) persuasive. The Examiner has supplemented the rejection using Fan with Park et al. (U.S. Patent Pre-Grant Publication No. 2022/0101984) as described above to address the newly amended claimed portion of “wherein the first direction of training and the second direction of training are used to adjust weights in layers of the machine learning model”. The 35 U.S.C. 103 rejection(s) stand.
In response to argument (5), the Examiner finds the Applicant’s argument(s) persuasive. The Examiner has supplemented the rejection using Fan with Park et al. (U.S. Patent Pre-Grant Publication No. 2022/0101984) as described above to address the current claims. The Examiner maintains that Fan when combined with Park encompasses claim 6. The 35 U.S.C. 103 rejection(s) stand.
In response to argument (6), the Examiner does not find the Applicant’s argument(s) persuasive. Fan describes in Paragraph [0116] that “Therefore, processor 1120, along with capture control module 1135, multi-aperture spectral camera 1160, and working memory 1105 represent one means for capturing a set of spectral images and/or a sequence of images.” which the Examiner maintains the means for capturing a set of spectral images and/or a sequence of images allows for multiple images to be present in the training (See Paragraph [0198]). The 35 U.S.C. 102 rejection(s) stand.
In response to argument (7), the Examiner does not find the Applicant’s argument(s) persuasive. The Examiner maintains that Fan teaches “the predicted metadata comprising one or more predicted wound properties of the wound at the corresponding future time” in Paragraphs [0160]-[0164]: “Wound assessment and/or healing predictions described herein may be accomplished based on one or more images of the wound, either alone or based on a combination of both patient health data (e.g., one or more health metric values, clinical features, etc.) and images of the wound, and patient metrics or a combination of some or all of the patient metrics to improve the accuracy of predicted healing parameters generated by the systems and methods”. The 35 U.S.C. 102 rejection(s) stand.
In response to argument (8), the Examiner maintains the 35 U.S.C. 102 rejection(s) and the 35 U.S.C. 103 rejection(s) on dependent claims 2-12, 14-20. The 35 U.S.C. 102 and 35 U.S.C. 103 rejection(s) stand.
Conclusion
THIS ACTION IS MADE FINAL. 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 nonprovisional extension fee (37 CFR 1.17(a)) 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Bennett S Erickson whose telephone number is (571)270-3690. The examiner can normally be reached Monday - Friday: 9:00am - 5:00pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Robert Morgan can be reached at (571) 272-6773. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Bennett Stephen Erickson/Primary Examiner, Art Unit 3683
Prosecution Insights