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 .
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1, 2, 5-6, 9-12, 14, and 21-24 rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood.
As to Claim 1, Alzamzmi teaches a non-transitory computer-readable medium having executable instructions stored thereon that, when executed by a processor, cause a system to perform operations comprising (see paragraph [0006], “The non-transitory computer readable medium comprising instructions that when executed by a computer, configure the computer to perform steps comprising”, where it is understood that the computer comprises a processor):
obtaining first video scan data comprising multiple first video frames of an inferior vena cava (IVC) of a subject (see paragraph [0005], “The method comprising: receiving images from the echocardiography study…localizing and segmenting the selected images to obtain IVC regions in the selected images”),
the multiple first video frames including a first video frame of the IVC while the subject is at rest, and a second video frame of the IVC while the subject is inhaling (see paragraph [0027], “the IVC Quantification and RAP estimation network 210 performs temporal analysis of cardiac indices over all frames thereby improving reproducibility in clinical cardiology practice and research . Analysis of all frames provides information about temporal changes during respiration over multiple cardiac cycles”, where respiration over multiple cardiac cycles would capture the subject inhaling and at rest);
and predicting, using a trained model, based at least on the multiple first video frames, a right atrial pressure (RAP) of the subject (see paragraph [0017], “Embodiments of the present disclosure provide an echocardiography artificial intelligence (AI) system that is capable of estimating inferior vena cava (IVC) collapsibility and RAP”, and see paragraph [0027], “Analysis of all frames provides information about temporal changes during respiration over multiple cardiac cycles”).
Alzamzmi fails to teach determining, using a first trained model, based at least on the multiple first video frames, that the first video scan data corresponds to a sniff test of the IVC of the subject. However, in an analogous art, Heywood teaches a machine learning model (see Col. 7, lines 46-52 “Preferably, the machine learning model(s) of detection process 247, policy compliance process 248, and/or interaction evaluation process 249 may include one or more deep-learning classifiers. Such a classifier may, for example, attempt to classify/label a given image or portion of an image based on a training set of labeled image data”),
which is trained to determine if video scan data corresponds to an inhalation (see Col. 10, lines 34-37, “In one embodiment, respiratory monitor 304 b may apply an image classifier to video data in sensor data 312, to detect inhalation and exhalation by the person”).
Thus, it would have been obvious one of ordinary skill the art before the effective filing date of the claimed invention to combine the pattern recognizer taught by Heywood with the RAP estimation taught by Alzamzmi. The motivation for doing so would be to alert a medical professional of the health condition of the patient. Heywood teaches in Col. 1, lines 52-57, “For instance, one alert may notify the first responder of the detected condition (e.g., to change the behavior of the first responder), while another alert (e.g., additionally and/or based on different thresholds) may dispatch emergency medical personnel, in the case of an individual exhibiting medical distress.” Thus, it would have been obvious to combine the teachings of Heywood with the RAP estimation technique taught by Alzamzmi in order to obtain the invention as claimed in Claim 1.
As to Claim 3, Alzamzmi in view of Heywood teaches prior to obtaining the first video scan data, obtaining second video scan data comprising multiple second video frames of the IVC (see Alzamzmi, paragraph [0021], “As illustrated in FIG. 2 , an echocardiography study 202 is performed where images are captured”);
and determining, using a third trained model, based at least on the multiple second video frames, that an acceptable view of the IVC is being captured (see Alzamzmi, paragraph [0023], “The view classification head detects an IVC view from a given echo study while the quality assessment head labels a given IVC view as good quality or bad quality”);
and obtaining the first video scan data comprises obtaining the first video scan data in response to determining that the acceptable view of the IVC is being captured (see Alzamzmi, paragraph [0022], “The image retrieval network 206 of the system 200 retrieves a specific view with acceptable (moderate to good) quality”).
As to Claim 5, Alzamzmi in view of Heywood teaches obtaining the first video scan data and the second video scan data comprises: capturing, using an ultrasound imaging device, a transthoracic echocardiogram of the subject (see Alzamzmi, paragraph [0003], “An echocardiogram (“echo”) is performed by using a dedicated bedside or portable imaging system to capture an ultrasound image of the heart and its associated anatomical structures”, and see Fig. 102 showing image from a transthoracic echocardiogram).
As to Claim 6, Alzamzmi in view of Heywood teaches obtaining multiple video scans, each of the multiple video scans corresponding to a medical imaging study (see Alzamzmi, paragraph [0023], “The image retrieval network 206 includes five inverted residual bottleneck (see MobileNetV2-s residual bottleneck [8]) blocks and a final pooling layer…The view classification head detects an IVC view from a given echo study while the quality assessment head labels a given IVC view as good quality or bad quality”, thus implying that scan data from multiple echo studies is obtained);
assigning multiple labels to the multiple video scans, each of the labels indicating whether or not a respective one of the video scans corresponds to a view of an IVC (see paragraph Alzamzmi, [0049], “In each run, the view classifier will either detect a specific view or label it as unknown. The clustering engine then groups unknown views into clusters (based on their similarity) to be labeled by a human expert before passing the newly labeled clusters/classes to the classification engine for model update”);
and training, based on the multiple video scans and the multiple labels, a classification model as the third trained model (see Alzamzmi, paragraph [0049], “For example, the view classification engine contains a classifier trained to recognize different echocardiography views including IVC view, but it is also trained using OpenMax (described further below) to recognize unknown views”).
As to Claim 9, Alzamzmi in view of Heywood teaches wherein predicting, using the second trained model, based at least on the multiple first video frames, the RAP of the subject comprises: inputting the multiple first video frames into the second trained model (see Alzamzmi, paragraph [0027], “Instead of extracting cardiac biomarkers in specific frames (e.g., end-diastolic and end-systolic) as is done manually, the IVC Quantification and RAP estimation network 210 performs temporal analysis of cardiac indices over all frames “;
and generating, using the second trained model, a prediction output including the RAP (see Alzamzmi, paragraph [0030], “the IVC Quantification and RAP estimation network 210 computes the RAP value”.
As to Claim 10, Alzamzmi in view of Heywood teaches wherein the RAP of the prediction output is between 0 mmHg and 30 mmHg (see Alzamzmi, paragraph [0030], “RAP is computed as follows: (1) the IVC Quantification and RAP estimation network 210 computes IVC collapsibility based on a difference between an absolute maximum peak and minimum valley in the IVCD curve; and (2) the IVC Quantification and RAP estimation network 210 computes the RAP value by plugging the IVC diameter and collapsibility values into equation (1)”, and see equation (1) shown below, where the output RAP is either 3, 8, or 15 mmHg, which are all values between 0 mmHg and 30 mmHg).
PNG
media_image1.png
138
601
media_image1.png
Greyscale
Equation (1) of Alzamzmi
As to Claim 11, Alzamzmi in view of Heywood teaches wherein the RAP of the prediction output is 3 mmHg, 8 mmHg, or 15 mmHg (see Alzamzmi, paragraph [0030], “RAP is computed as follows: (1) the IVC Quantification and RAP estimation network 210 computes IVC collapsibility based on a difference between an absolute maximum peak and minimum valley in the IVCD curve; and (2) the IVC Quantification and RAP estimation network 210 computes the RAP value by plugging the IVC diameter and collapsibility values into equation (1)”, and see Equation (1) shown above, where the output RAP output is either 3, 8, or 15 mmHg).
As to Claim 12, Alzamzmi in view of Heywood teaches obtaining each of the multiple video scans corresponding to a sniff test of a subject,
obtaining multiple video scans; obtaining multiple RAP measurements, each of the RAP measurements corresponding to a respective one of the video scans (see paragraph [0034], “To assess the efficacy of the RAP estimation performed by the IVC Quantification and RAP estimation network 210, the automated RAP values are compared against those estimated by experts”);
and constructing, based on the multiple video scans and the multiple RAP measurements, the second trained model (see paragraph [0049], “In the last stage of the system, an open-world active learning approach is integrated to the disease classification model, which uses the automated IVC diameter and collapsibility. This disease classification model will classify known cardiac diseases into their respective classes and identify new (unseen) diseases as unknown”).
As to Claim 14, Alzamzmi in view of Heywood teaches wherein each RAP measurement of the RAP measurements is an RAP estimate made by a physician based on the respective one of the video scans corresponding to the RAP measurement (see Alzamzmi, paragraph [0034], “To assess the efficacy of the RAP estimation performed by the IVC Quantification and RAP estimation network 210, the automated RAP values are compared against those estimated by experts”, wherein it is understood that the ‘expert’ is a physician).
As to Claim 21, Alzamzmi in view of Heywood teaches a system, comprising: a processor; and a non-transitory computer-readable medium having executable instructions stored thereon that, when executed by the processor (see Alzamzmi, paragraph [0006], “The non-transitory computer readable medium comprising instructions that when executed by a computer, configure the computer to perform steps comprising”, and it is understood that a computer must inherently contain a processor), cause the system to perform operations comprising the same steps recited in Claim 1. Therefore, the rejection and rationale are analogous to that of Claim 1.
As to Claim 22, Alzamzmi in view of Heywood teaches using an ultrasonic imaging device configured to capture the first video scan data (see Alzamzmi, paragraph [0003], “An echocardiogram (“echo”) is performed by using a dedicated bedside or portable imaging system to capture an ultrasound image of the heart and its associated anatomical structures”, where it is understood that the imaging system capturing the ultrasound system is an ultrasonic imaging device).
As to Claim 23, Alzamzmi in view of Heywood teaches the ultrasonic imaging device is a portable ultrasonic imaging device (see Alzamzmi, paragraph [0003], “An echocardiogram (“echo”) is performed by using a dedicated bedside or portable imaging system to capture an ultrasound image of the heart and its associated anatomical structures”)..
As to Claim 27, Alzamzmi in view of Heywood teaches a method (see Alzamzmi, paragraph [0005], “In another aspect of the disclosure, a method for estimating inferior vena cava (IVC) collapsibility and right atrial pressure (RAP) from an echocardiography study in real-time performed by an Artificial Intelligence (AI) system is provided”), comprising the same steps recited in Claim 1. Therefore, the rejection and rationale are analogous to that of Claim 1.
Claim(s) 4 is rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood and further in view of Arnaout (US Pub No 20220012875), hereinafter Arnaout.
As to Claim 4, Alzamzmi in view of Heywood teaches a model that may determine an acceptable view of the IVC is being captures based at least on the multiple video frames, (see Alzamzmi, paragraph [0022], “The retrieval is performed with a lightweight model with a shared encoder and two heads (see FIG. 7 and related discussion below for further details), with the first head for view classification to identify an IVC view and the second head for quality assessment to assess the quality of the identified view”); However Alzamzmi fails to teach the train model generates a prediction comprising a confidence score that indicates a likelihood that the second video scan data is associated with an IVC class; and in response to determining that the confidence score meets a threshold, makes a determination that an acceptable view of the IVC is being captured.
However, in an analogous art, Arnaout teaches a machine learning model for determining a quality of a medical image (see paragraph [0042], “In certain embodiments, image analysis engines employ a variety of machine learning techniques to determine a quality level of images captured by a capture element for use in diagnosis”),
which comprises obtaining ultrasound images of a heart (see paragraph [0042], “In various embodiments, a system for medical image capture can be used to capture and evaluate ultrasound images, such as echocardiograms”),
generating, using the third trained model prediction comprising a confidence score that indicates a likelihood that the second video scan data is associated with a class (see paragraph [0048], “Process 100 identifies (115) target view images from the input images. In certain embodiments, processes can perform a classification operation on the target view images to calculate a probability of the image belonging to each of several possible classes”);
and in response to determining that the confidence score meets a threshold, making a determination that an acceptable view is being captured (see paragraph [0022], “evaluating image frames includes determining whether image frames of a quality level greater than a threshold value have been captured for each of the several target views, wherein the quality level is based on a confidence level for a classification”).
Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the prediction score and threshold taught by Arnaout with the machine learning model for IVC view classification taught by Alzamzmi. The motivation for doing so would be to ensure the correct view is obtained. Arnaout teaches in paragraph [0065], “Particularly in the case of abnormal hearts, it can often be difficult to ensure that an ultrasound contains quality images for each of the target views. Processes in accordance with various embodiments of the invention can provide instructions (e.g., visual guidance on a recommended angle and/or location for directing a user) to assist in the capture of a desired view. In many embodiments, processes can only provide feedback once a video capture has been completed, or periodically during an image capture. By selecting images with scores exceeding a threshold, the specificity for views of interest can be increased.”. Thus, it would have been obvious to combine the prediction scores taught by Arnaout with the teachings of Alzamzmi and Heywood in order to obtain the invention as claimed in Claim 4.
Claim 13 is rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood, and further in view of Zamzmi et al. (Zamzmi G., et al. “Evaluation of an artificial intelligence-based system for echocardiographic estimation of right atrial pressure”, Int J Cardiovasc Imaging. 2023), hereinafter Zamzmi.
As to Claim 13, Alzamzmi in view of Heywood fails to explicitly teach the multiple video scans were captured by a plurality of different models of ultrasound imaging machines; and the multiple RAP measurements comprise a plurality of RAP estimates made by a plurality of different cardiologists.
However, in an analogous art, Zamzmi teaches a method of obtaining a right atrial pressure through machine learning (see pg. 2437, Abstract, “This study aims to develop a fully automated artificial intelligence (AI)-based system for automated IVC analysis and RAP estimation”),
wherein multiple video scans were captured by a plurality of different models of ultrasound imaging machines (see pg. 2439, ‘Echocardiography dataset’, “These were acquired using diverse echocardiography devices including iE33, GE E9, and GE Vivid E95”, wherein the listed devices are ultrasound machines),
and the multiple RAP measurements comprise a plurality of RAP estimates made by a plurality of different cardiologists (see pg. 2439, ‘Echocardiography dataset’ , “The manual measurements of IVC diameter were provided by board-certified echocardiographers following conventional methodology in current clinical practice” and see pg. 2441, ‘Statistical Analysis’, “To assess the agreement between the manual and automated RAP measurements,”).
Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the video scan data taught by Zamzmi with the teachings of Alzamzmi and Heywood. The motivation for doing so would be to increase the robustness of the model. Zamzmi teaches on pg. 2438, “The open-world feature makes the system more robust in detecting and learning new, unpredictable cases or scenarios in real-world clinical settings”. Thus , it would have been obvious to combine the video scan data taught by Zamzmi with the teachings of Alzamzmi and Heywood in order to obtain the invention as claimed in Claim 13.
Claims 15-17 are rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood, and further in view of Albani et al. (Albani, S., et al., “Accuracy of right atrial pressure estimation using a multi-parameter approach derived from inferior vena cava semi-automated edge-tracking echocardiography: a pilot study in patients with cardiovascular disorders.”, Int J Cardiovasc Imaging 36, 1213–1225 (2020)), hereinafter Albani.
As to Claim 15, Alzamzmi in view of Heywood fails to teach wherein each RAP measurement of the RAP measurements is a right heart catheterization (RHC) measurement made via RHC of a subject.
However, in an analogous art, Albani teaches an automated system for calculating RAP (see Abstract, pg. 1213, “The aim of this study is to assess feasibility and accuracy of a new semi-automated approach to estimate RAP”)
In which RAP measurements are obtained via right heart catheterization (RHC) (see Abstract, pg. 1213, “Direct RAP measurements obtained during a right heart catheterization (RHC) were used as reference.”). Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the RHC measurement taught by Albani with the machine learning model taught by Alzamzmi in view of Heywood. The motivation for doing so would be to use RHC as a reference to verify the accuracy of the mode. Albani teaches on pg. 1214, “The aim of this study is to assess the accuracy of the estimation of RAP using two different approaches based on the semi-automated tracking technique [9, 11], compared to standard echocardiographic methods. Direct RAP measurement obtained during a RHC was used as reference.” Thus, it would have been obvious to combine the teaching so Albani with the teachings of Alzamzmi in view of Heywood in order to obtain the invention as claimed in Claim 15.
As to Claim 16, Alzamzmi in view of Heywood fails to teach wherein each of the RHC measurements corresponds to a respective one of the video scans made of a same subject within one month or less.
However, Albani teaches each of the RHC measurements corresponds to a respective one of the video scans made of a same subject within one month or less. (see pg. 1214, “Exclusion criteria were: age < 18 years, more than 6 h between the invasive and echocardiographic assessment and liquid assumption or diuretics administration between the invasive and the ultrasonography assessments”). Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the RHC measurement taught by Albani with the machine learning model taught by Alzamzmi in view of Heywood. The motivation for doing so would be to use RHC as a reference to verify the accuracy of the mode (see Albani, pg. 1214). Thus, it would have been obvious to combine the teaching so Albani with the teachings of Alzamzmi in view of Heywood in order to obtain the invention as claimed in Claim 15.
As to Claim 17, Alzamzmi in view of Heywood teaches the multiple video scans comprise a first plurality of video scans (see Alzamzmi, paragraph [0004], “The system comprising: an image retrieval network configured to receive images from the echocardiography study”), and multiple RAP measurements comprising a plurality of RAP measurements, where each of the plurality of RAP measurements is an RAP estimate made by a physician based on a respective one of the first plurality of video scans (see paragraph [0034], “To assess the efficacy of the RAP estimation performed by the IVC Quantification and RAP estimation network 210, the automated RAP values are compared against those estimated by experts”).
Alzamzmi in view of Heywood fails to teach the multiple video scans first plurality of video scans and a second plurality of video scans; and the multiple RAP measurements comprise a first plurality of RAP measurements and a second plurality of RAP measurements and each of the second plurality of RAP measurements is a RHC measurement made via RHC of a subject, and corresponds to a respective one of the second plurality of video scans made of a same subject.
However, Albani teaches obtaining RAP measurements via a RHC of a subject, where each RHC corresponds to a plurality of video scan data (see pg. 1214, “The aim of this study is to assess the accuracy of the estimation of RAP using two different approaches based on the semi-automated tracking technique [9, 11], compared to standard echocardiographic methods. Direct RAP measurement obtained during a RHC was used as reference”, and see pg. 1215, “A semi-automated algorithm was used to process US video clips”, where US stands for ultrasound).
Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the RHC measurement and corresponding RAP dataset taught by Albani with the RAP dataset taught by Alzamzmi in view of Heywood. The motivation for doing so would be to use RHC as a reference to verify the accuracy of the mode (see Albani, pg. 1214). Thus, it would have been obvious to combine the teaching so Albani with the teachings of Alzamzmi in view of Heywood in order to obtain the combined RAP dataset as claimed in Claim 15.
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood, in view of Albani et al. (Albani, S., et al., “Accuracy of right atrial pressure estimation using a multi-parameter approach derived from inferior vena cava semi-automated edge-tracking echocardiography: a pilot study in patients with cardiovascular disorders.”, Int J Cardiovasc Imaging 36, 1213–1225 (2020)), hereinafter Albani, and further in view of Honkala et al. (US Pub No 20190340754), hereinafter Honkala.
As to Claim 18, Alzamzmi in view of Heywood and Albani teaches a first plurality of video scans and a first plurality of RAP measurements (see Alzamzmi, paragraph [0004], “The system comprising: an image retrieval network configured to receive images from the echocardiography study”), and see paragraph [0034], “To assess the efficacy of the RAP estimation performed by the IVC Quantification and RAP estimation network 210, the automated RAP values are compared against those estimated by experts”),
and a second plurality of video scans and a second plurality of RAP measurements (see Albani, pg. 1214, ““The aim of this study is to assess the accuracy of the estimation of RAP using two different approaches based on the semi-automated tracking technique [9, 11], compared to standard echocardiographic methods. Direct RAP measurement obtained during a RHC was used as reference.”, and see pg. 1215, “A semi-automated algorithm was used to process US video clips” ) wherein constructing, based on the multiple video scans and the multiple RAP measurements, the second trained model, comprises: pre-training, based on the first plurality of video scans and the first plurality of RAP measurements, an input model to estimate RAP based on an input video scan ( see paragraph [0030], “the IVC Quantification and RAP estimation network 210 computes the RAP value”, and see paragraph [0049], “In the last stage of the system, an open-world active learning approach is integrated to the disease classification model, which uses the automated IVC diameter and collapsibility. This disease classification model will classify known cardiac diseases into their respective classes and identify new (unseen) diseases as unknown”).
Alzamzmi in view of Heywood and Albani fails to teach applying, based on the second plurality of video scans and the second plurality of RAP measurements, transfer learning to the input model to construct the second trained model.
However, in an analogous art, Honkala teaches a method for analyzing medical images (see paragraph [0037], “The first and second subsequent images may be X-ray image…positron emission tomography (PET) images, single photon emission tomography (SPET) images, Magnetoencephalography (MEG) images or ultrasound images”
where transfer learning can be used to train a model on a new dataset (see paragraph [0118], “For example, in transfer learning, a pre-trained model can be re-trained with new data to improve performance or to perform a new task. In multi-task learning, multiple criteria (such as denoising and diagnosis) may be used simultaneously during training. Transfer learning and/or multi-task learning may be used to improve the training and/or the accuracy of the resulting model”).
Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the transfer learning taught by Honkala with the first videos and RAP values taught by Alzamzmi and the second videos and RAP values taught by Albani. The motivation for doing so would be to improve performance of the model, as taught by Honkala in paragraph [0118]. Thus, it would have been obvious to combine the transfer learning taught by Honkala with the teachings of Alzamzmi, Heywood, and Albani in order to obtain the invention as claimed in Claim 18.
Claims 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood, in view of Albani et al. (Albani, S., et al., “Accuracy of right atrial pressure estimation using a multi-parameter approach derived from inferior vena cava semi-automated edge-tracking echocardiography: a pilot study in patients with cardiovascular disorders.”, Int J Cardiovasc Imaging 36, 1213–1225 (2020)), hereinafter Albani, further in view of Honkala et al. (US Pub No 20190340754), hereinafter Honkala, and further in view of Kopparapu (US Pub 20190156159), hereinafter Kopparapu.
As to Claim 19, Alzamzmi in view of Heywood, Albani, and Honkala fails to teach wherein applying transfer learning to the input model to construct the second trained model comprises: replacing an output layer of the input model with a new output layer to produce a new model; and training, using the second plurality of video scans and the second plurality of RAP measurements, the new model.
However, in an analogous art, Kopparapu teaches applying transfer learning to an input model (see paragraph [0112], “Specifically, the disclosure relates to a platform that uses trained machine learning architecture, for example, neural networks…networks initialized via a transfer learning technique”),
which comprises replacing an output layer of the input model with a new output layer to produce a new model (see paragraph [0151], “In order to generate the pixel-wise classification, the CNN is modified to an FCN in two steps. First, the final three fully-connected layers are eliminated and replaced with convolutional layers”);
and re-training in order to obtain a new model (see paragraph [0150], “Following the transplantation of the trained AlexNet weights into modified FCN-AlexNet, the network is fine-tuned by continuing the backpropagation process for about 20 epochs of training”).
Thus, it would have been obvious to one or ordinary skill in the art before the effective filing date of the claimed invention to combine the layer replacement and transfer learning taught by Kopparapu with the teachings of Alzamzmi, Heywood, Albani, and Honkala. The motivation for doing so would be to allow the model to process larger images. Kopparapu teaches in paragraph [0151], “In order to generate the pixel-wise classification, the CNN is modified to an FCN in two steps. First, the final three fully-connected layers are eliminated and replaced with convolutional layers. This change allows the network to process larger images (as opposed to the original 224×224 pixels) by sliding a window across the image and generating a classification per window.” Thus, it would have been obvious to combine the layer replacement taught by Kopparapu with the teachings of Alzamzmi, Heywood, Albani, and Honkala in order to obtain the invention as claimed in Claim 19.
As to Claim 20, Alzamzmi in view of Heywood, Albani, and Honkala fails to teach wherein training the new model comprises assigning a lower learning rate to pre-existing layers of the input model present in the new model compared to the new output layer.
However Kopparapu teaches training the new model comprises assigning a lower learning rate to pre-existing layers of the input model present in the new model compared to the new output layer (see paragraph [0152], “Hyperparameters for the training process are chosen based on the future of the transfer learning technique. A low base learning rate of 0.001 may be chosen to avoid greatly distorting the pre-trained weights”).
Thus it would have been obvious to combine the transfer learning and low learning rate taught by Kopparapu with the teachings of Alzamzmi, Heywood, Albani, and Honkala. The motivation for doing so would be to prevent the distortion of pre-trained weights (see Kopparapu, paragraph [0152]). Thus, it would have been obvious to combine the learning rate taught by Kopparapu with the teachings of Alzamzmi, Heywood, Albani, and Honkala in order to obtain the invention as claimed in Claim 19.
Claim(s) 24-25 are rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood, in view of Wetterling et al. (US Pub No 20250134496), hereinafter Wetterling.
As to Claim 24, Alzamzmi in view of Heywood fails to explicitly teach the processor and non-transitory computer-readable medium are components of a mobile device; and the mobile device is communicatively coupled to the portable ultrasonic imaging device.
However, Wetterling teaches a non-transitory computer readable medium processor and (see paragraph [0056], “The invention further provides a computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out the aforementioned method” which is coupled to a mobile ultrasound imaging device (see paragraph [0058], “The ultrasound images may be provided by a portable ultrasound device comprising: an ultrasound transducer comprising a 2D array of independently controllable ultrasound transducer elements for producing an ultrasonic pulse…and analysis means”, wherein the analysis means comprises the processor and computer readable medium).
Thus, it would have been obvious to one of ordinary skill in the art to modify the system taught by Alzamzmi and Heywood so that it is portable, as taught by Wetterling. The motivation for doing so would be so that the system can easily be brought to the patient. Alzamzmi teaches in paragraph [0003], “An echocardiogram (“echo”) is performed by using a dedicated bedside or portable imaging system to capture an ultrasound image of the heart and its associated anatomical structures”). Thus, one would have been motivated to combine the teachings of Wetterling with the teachings of Alzamzmi and Heywood in order to obtain the invention as claimed in Claim 24.
As to Claim 25, Alzamzmi in view of Heywood fails to explicitly wherein the system is a mobile device. However, Wetterling teaches that the system is a mobile device (see paragraph [0058], “The ultrasound images may be provided by a portable ultrasound device comprising: an ultrasound transducer comprising a 2D array of independently controllable ultrasound transducer elements for producing an ultrasonic pulse…and analysis means”, wherein the analysis means comprises the processor and computer readable medium). Thus, it would have been obvious to one of ordinary skill in the art to modify the system taught by Alzamzmi and Heywood so that it is portable, as taught by Wetterling. The motivation for doing so would be so that the system can easily be brought to the patient. Thus, one would have been motivated to combine the teachings of Wetterling with the teachings of Alzamzmi and Heywood in order to obtain the invention as claimed in Claim 25.
Claim 26 is rejected under 35 U.S.C. 103 as being unpatentable over Alzamzmi et al. (US Pub No 20250315947), hereinafter Alzamzmi, in view Heywood et al. (US Pat No 11706391) of hereinafter, hereinafter Heywood, in view of Minor et al. (US Pub No 20200196944 ), hereinafter Minor.
As to Claim 26, Alzamzmi in view of Heywood fails to explicitly wherein the operations further comprise: displaying, on a graphical user interface, the RAP that is predicted.
However, Minor teaches a system comprising a graphical user interface (see paragraph [0041], “a monitoring system, comprising: a receiver configured to receive measurements associated with the left heart pressure and the right heart pressure; a memory unit configured to store the received measurements; a display device”)
wherein the RAP can be displayed (see abstract, “The medical system further comprises a receiver configured to receive the measurement data corresponding to the sensed right atrial pressure and the sensed left atrial pressure, and output to a display device the received measurement data”).
Thus, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the display taught by Minor with the system taught by Alzamzmi in view of Heywood. The motivation for doing so would be to. Thus, it would have been obvious to combine the display taught by Minor with the teachings of Alzamzmi and Heywood in order to obtain the invention as claimed in Claim 26.
Allowable Subject Matter
Claims 2, 7, and 8 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
The following is a statement of reasons for the indication of allowable subject matter:
As to Claim 2, all of the previously cited art fail to teach “wherein predicting the RAP of the subject is performed in response to determining that the first video scan data corresponds to a sniff test of the IVC of the subject”. Although Alzamzmi teaches predicting a RAP value (see paragraph [0004]), Alzamzmi fails to explicitly teach determining first video scan data corresponds to a sniff test using a machine learning model. Heywood teaches a classifier which can be used to determine if video data corresponds to patient breathing (see Col. 10, lines 34-37), but fails to teach using this video data to determine or predict a RAP.
In an analogous art, Wetterling et al. (WO Pub No 2023170632), teaches a machine learning model which can determine if data corresponds to a sniff test, (see paragraph [0023], “In one embodiment, pattern recognizer 310 compares incoming area trace signals with known area trace patterns to determine whether the incoming area trace is reflective of a feature such as a signal response to a patient maneuver. “Maneuver” as used herein refers to a physical action taken by a patient, on his or her own initiative or in response to instructions, which stimulates an identifiable perturbation of IVC area. Some examples of area traces for different patient maneuvers are shown in FIG. 6, including supine: quiet respiration, sniff, supine”). Wetterling further teaches that RAP may be obtained (see paragraph [0047], “Feature outputs from database 510 (or directly from upstream processes), invoke process 522 to generate metrics as described herein, including, for example, Congestion Index 524, Right Atrial Pressure (RAP) 526”. However, the data analyzed by Wetterling et al. is ‘trace data’, not video frame data as required by Claim 1. Furthermore, Wetterling et al. does not explicitly teach that the RAP is predicted in response to determining if the data contains ‘a sniff’.
Vaidya (US Pub No 20220304654 ) teaches a machine learning model which can be used to determine IVC volume during deep breathing or sniff. However, Vaidya fails to explicitly teach that the video data is analyzed to determine if the video data obtained corresponds to a ‘sniff test’.
As to Claim 7, all of the previously cited art fail to teach “generating, using the first trained model, based at least on the multiple first video frames, a prediction comprising a confidence score that indicates a likelihood that the first video scan data is associated with a sniff test class and in response to determining that the confidence score meets a threshold, making a determination that the first video scan data corresponds to the sniff test of the IVC of the subject”.
Heywood teaches outputting a confidence score related to video data. However, this confidence score does not indicate a likelihood that the first video scan data is associated with a sniff test class, instead, the confidence score is related to overall patient health, not whether the video data indicates the inhalation of a user (see Col. 12, lines 27-35, “In various embodiments, alert generator 308 may take as input the outputs of audio analyzer 302, health analyzer 304, and video analyzer 306, to determine whether an alert 314 should be issued. In some embodiments, such outputs may take the form of classification labels and associated probability/confidence measures, estimated values (e.g., pulse rate, respiratory rate, etc.), or the like”).
As to Claim 8, all of the previously cited art fail to teach obtaining multiple video scans, each of the multiple video scans corresponding to a medical imaging study; assigning multiple labels to the multiple video scans, each of the labels indicating whether or not a respective one of the video scans corresponds to a view of an IVC during a sniff test; and training, based on the multiple video scans and the multiple labels, a classification model as the first trained model.
Heywood teaches obtaining labelled data, and training a model based on the labelled data (see Col. 8, lines 15-35). However, Heywood fails to explicitly teach training a model to recognize patients inhalation through obtaining labelled datasets.
In an analogous art, Wetterling et al. (WO Pub No 2023170632), teaches a machine learning model which can determine if data corresponds to a sniff test, (see paragraph [0023]), which is trained on supervised data (see paragraph [0024], “For example, quality checks could be trained on supervised data as, for instance, the type of maneuver prescribed.”). However, Wetterling et al. does not teach using video data to determine if a sniff has occurred; instead, ‘trace data’ obtained from an ultrasound is used.
Conclusion
Wetterling et al. (WO Pub No 2023170632), teaches a machine learning model which can determine if trace data obtained from a patient ultrasound corresponds to a sniff test. Wetterling further teaches that RAP prediction may be obtained.
Vaidya (US Pub No 20220304654) teaches a machine learning model which can be used to determine IVC volume during deep breathing or sniff.
Sweeney et al. (WO Pub No 2024180503) teaches a method for monitoring the venous pressure of a patient which comprises obtaining ultrasound data during a patient maneuver such as a sniff.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SOUMYA THOMAS whose telephone number is (571)272-8639. The examiner can normally be reached M-F 8:30-5:00.
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, Jennifer Mehmood can be reached at (571) 272-2976. 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.
/S.T./Examiner, Art Unit 2664
/JENNIFER MEHMOOD/Supervisory Patent Examiner, Art Unit 2664
Prosecution Insights