Prosecution Insights
Last updated: November 20, 2026
Application No. 18/870,024

ARTICLES AND METHODS FOR FORMAT INDEPENDENT DETECTION OF HIDDEN CARDIOVASCULAR DISEASE FROM PRINTED ELECTROCARDIOGRAPHIC IMAGES USING DEEP LEARNING

Non-Final OA §101§102§103§112
Filed
Nov 27, 2024
Priority
May 27, 2022 — provisional 63/346,610 +1 more
Examiner
MOTSINGER, SEAN T
Art Unit
2673
Tech Center
2600 — Communications
Assignee
Yale University
OA Round
1 (Non-Final)
79%
Grant Probability
Favorable
1-2
OA Rounds
11m
Est. Remaining
90%
With Interview

Examiner Intelligence

Grants 79% — above average
79%
Career Allowance Rate
552 granted / 702 resolved
+16.6% vs TC avg
Moderate +12% lift
Without
With
+11.8%
Interview Lift
resolved cases with interview
Typical timeline
2y 11m
Avg Prosecution
23 currently pending
Career history
736
Total Applications
across all art units

Statute-Specific Performance

§101
14.3%
-25.7% vs TC avg
§103
40.1%
+0.1% vs TC avg
§102
19.2%
-20.8% vs TC avg
§112
19.3%
-20.7% vs TC avg
Black line = Tech Center average estimate • Based on career data from 702 resolved cases

Office Action

§101 §102 §103 §112
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 § 101 35 U.S.C. 101 reads as follows: Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title. Claim 1-23 rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because Re claim 1 The limitation of “receiving an electrocardiogram (ECG) image for the subject”, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, receiving an electrocardiogram in the context of this claim encompasses a user looking at an electrocardiogram to receive the image. Similarly, the limitation of applying a algorithm to the ECG image for the subject, the algorithm being trained to distinguish a printed ECG reading of a heart with cardiovascular disease from a printed ECG reading of a healthy heart, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, an algorithm in the context of this claim encompasses applying a mental algorithm. Similarly, the limitation of comparing outputs of the algorithm to patterns of algorithm outputs for ECG images from healthy subjects and subjects with one or more cardiovascular diseases, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, comparing in the context of this claim encompasses a user mentally from performing the comparison. Similarly, the limitation of determining if the subject has cardiovascular disease based upon the outputs of the algorithm, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, determining in the context of this claim encompasses a user mentally performing the comparison. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. This judicial exception is not integrated into a practical application. In particular, the claim only recites additional elements – using a computer to implement the method and algorithm is a machine-learning based algorithm. The computer in the method is recited at a high-level of generality (i.e., as a generic computer performing a generic computer function ) such that it amounts no more than mere instructions to apply the exception using a generic computer component. The machine learning based limitation merely limits the claim to the field of machine learning. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a computer amounts to no more than mere instructions to apply the exception using a generic computer component. Furthermore, the machine learning based limitation merely limits the claim to the field of machine learning. Mere instructions to apply an exception using a generic computer component combined with the generic “machine learning“ cannot provide an inventive concept. The claim is not patent eligible. Re claim 2 The limitation of algorithm is trained to distinguish a printed ECG reading of a heart with cardiovascular disease from a printed ECG reading of a healthy heart, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, the algorithm in the context of this claim encompasses a mental algorithm for distinguishing. This judicial exception is not integrated into a practical application. In particular, the claim only recites additional elements – using a computer to implement the method and algorithm is a deep neural network, the deep neural network comprising a plurality of nodes. The computer in the method is recited at a high-level of generality (i.e., as a generic computer performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer component. The deep learning element merely describes a generic deep learning algorithm described at high level. The deep learning limitation merely limits the claim to the generic field of deep learning. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a computer amounts to no more than mere instructions to apply the exception using a generic computer component. Furthermore, the deep learning limitation merely limits the claim to the field of deep learning. Mere instructions to apply an exception using a generic computer component combined with the generic “deep learning” cannot provide an inventive concept. The claim is not patent eligible. Re claim 3 The limitation of the algorithm is a statistical algorithm, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, algorithm in the context of this claim encompasses a mental algorithm. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 4 The limitation of wherein the ECG image comprises a printed ECG image of an ECG dataset formed by conversion of ECG waveform data, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, receiving a printed image in the context of this claim encompasses a user mentally looking at a printed image to receive it. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 5 The limitation of wherein the method is generalizable to multiple ECG image formats, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, the method in the context of this claim encompasses a mental method which is generalizable. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 6 The limitation of algorithm trained on ECG images having incorrectly placed leads, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, training in the context of this claim encompasses a user mentally training a mental algorithm on images with incorrectly placed leads. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 7 The limitation of algorithm is trained on images of ECGs with different signal, background, and noise characteristics, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, the algorithm in the context of this claim encompasses a mental algorithm which has been mentally trained. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 8 the limitation of identifying hidden clinical labels, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, identifying hidden labels in the context of this claim encompasses mentally identifying a label. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 9 The limitation of identifying characteristics of the ECG image that the determination is based on, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, identifying in the context of this claim encompasses a user mentally identifying the characteristics. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 10 claim 10 contains the same abstract idea as claim 1. This judicial exception is not integrated into a practical application. In particular, the claim only recites additional elements – using a computer to automate the method and the algorithm is a machine-learning based algorithm. The computer in the method is recited at a high-level of generality (i.e., as a generic computer performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer component. The machine learning based limitation merely limits the claim to the field of machine learning. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a computer amounts to no more than mere instructions to apply the exception using a generic computer component. Furthermore, the machine learning based limitation merely limits the claim to the field of machine learning. Mere instructions to apply an exception using a generic computer component combined with the generic “machine learning“ cannot provide an inventive concept. The claim is not patent eligible. Re claim 11 the limitation of wherein the cardiovascular disease comprises a disorder selected from the group consisting of structural disorders of the heart, functional disorders of the heart, structural disorders of the structures supporting the heart, functional disorders of the structures supporting the heart, and combinations thereof, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders the algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 12 The limitation of wherein the cardiovascular disease comprises a disorder selected from the group consisting of structural disorders of the heart, functional disorders of the heart, structural disorders of the structures supporting the heart, functional disorders of the structures supporting the heart, and combinations thereof, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 13 The limitation of wherein the disorder is a genetic disorder, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 14 The limitation of wherein the disorder is an acquired disorder, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 15 The limitation of wherein the cardiovascular disease comprises a disease that is not normally discernable by physicians from ECG data, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders the mental algorithm. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 16 the limitation of wherein, prior to the step of applying the algorithm to the ECG image for the subject, the method further comprises training the algorithm, the training of the algorithm comprising: creating an image-based dataset including a normal subset and a cardiovascular disease subset; optionally pre-training the algorithm on an unrelated clinical or hidden label; and training the algorithm on the image-based dataset, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, creating a data set and training in the context of this claim encompasses mentally determining a data set and mentally training a mental model. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 17 The limitation of wherein the cardiovascular disease subset includes a low ejection fraction (EF) subset as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders of mental data set the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 18 The limitation of wherein the low EF subset includes ECG images for individuals with EF of less than 40% as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders of mental data set the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 19 the limitation of wherein the clinical label includes six physician-defined labels and the hidden label includes gender, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, the labels in the context of this claim encompasses a user mentally identifying 6 clinical labels. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 20 The limitation of wherein the normal subset includes ECG images for individuals having hypertrophic cardiomyopathy (HCM), as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders of mental data set the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 21 The limitation of wherein the cardiovascular disease subset includes ECG images for individuals having HCM and left ventricular (LV) systolic dysfunction, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders of mental data set the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 22 The limitation of wherein the image-based dataset includes at least two different plotting schemes for each ECG waveform, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders of mental data set the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Re claim 23 The limitation of wherein the image-based dataset includes at least two different ECG image formats, as drafted is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, in the context of this claim this element merely further defines the disorders of mental data set the mental algorithm acts on. The analysis with respect to integration into an abstract idea and significantly more is not substantially changed from the claim from which this claim depends. Claim Rejections - 35 USC § 112 The following is a quotation of 35 U.S.C. 112(b): (b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention. The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph: The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention. Claims 16-23 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. Re claim 16 The claim recites “optionally pre-training the algorithm on an unrelated clinical or hidden label”. The examiner notes that the word optionally is indefinite because it creates confusion as to whether or not the claim is actually required. Claims 17-23 depend from claim 16 and retain the same issue. Claim Rejections - 35 USC § 102 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 – Claim(s) 1-12, 16, 19 22-23 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Sangha et al “Automated multilabel diagnosis on electrocardiographic images and signals” Nature Communications volume 13, Article number: 1583 (2022) March 24 2022. Re claim 1 Sangha discloses A computer-implemented method of detecting cardiovascular disease in a subject, the method comprising: receiving an electrocardiogram (ECG) image for the subject (see page 6 discussion section first paragraph “We have developed an externally validated multilabel automated diagnosis algorithm that accurately identifies rhythm and conduction disorders from either ECG images” note that ECG images are used); applying a machine-learning based algorithm to the ECG image for the subject, the algorithm being trained to distinguish a printed ECG reading of a heart with cardiovascular disease from a printed ECG reading of a healthy heart (see method section model overview part “For image data, we built a CNN model based on the EfficientNet set of architectures10. To balance the complexity and accuracy of a model with the computational resources required for training, we chose the B3 architecture, which required PNG images to be sampled at 300 × 300 square pixels. The model includes 384 layers, has over 10 million trainable parameters, and is composed of seven blocks (Supplementary Fig. 1). Additional dropout layers were added to prevent overfitting, and the output of the EfficientNet-B3 model was fed into a fully connected layer with sigmoid activation to predict the probability of target labels.” Machine learning is used to determine if the ECG has any of 6 rhythm disorder labels See also Results section Study Population part note that only some of the ECGs used has a rhythm disorder see also “Performance of image-based multilabel classification” section second paragraph note that the training set only some of ECG had clinical labels); comparing outputs of the algorithm to patterns of algorithm outputs for ECG images from healthy subjects and subjects with one or more cardiovascular diseases( see method section model overview part “this information was passed through two more fully connected layers, the second of which had sigmoid activation to predict the probability of target labels” note that outputs are compared with an activation layer to predicted the labels as to whether or not the ECG has any of the conditions, note that as discussed above the model is trained to differentiate between Normal ECG and those with a clinical label); and determining if the subject has cardiovascular disease based upon the outputs of the algorithm (see method section model overview part “this information was passed through two more fully connected layers, the second of which had sigmoid activation to predict the probability of target labels” note that whether or not an ECG has any of a plurality of labels are determined.). Re claim 2 Sangha discloses wherein the machine-learning based algorithm is a deep neural network, the deep neural network comprising a plurality of nodes (see “Performance of image-based multilabel classification“ 5th paragraph “For image data, we built a convolutional neural network (CNN) model based on the EfficientNet B3 architecture (see “Methods” for details)10. We trained our primary image model on both the standardized subset, with a 40%-40%-10%-10% split of standard, alternate, standard shuffled, and alternate shuffled images as well as the real-world variation subset” note that a convolution al neural network is used, note that neural networks that the Effcient net could be considered deep learning and neural networks have nodes) trained to distinguish a printed ECG (see “” first paragraph “The ECGs were transformed to the corresponding printed images using the python library ecg-plot”) reading of a heart with cardiovascular disease from a printed ECG reading of a healthy heart (see method section model overview part “For image data, we built a CNN model based on the EfficientNet set of architectures10. To balance the complexity and accuracy of a model with the computational resources required for training, we chose the B3 architecture, which required PNG images to be sampled at 300 × 300 square pixels. The model includes 384 layers, has over 10 million trainable parameters, and is composed of seven blocks (Supplementary Fig. 1). Additional dropout layers were added to prevent overfitting, and the output of the EfficientNet-B3 model was fed into a fully connected layer with sigmoid activation to predict the probability of target labels.” Machine learning is used to determine if the ECG has any of 6 rhythm disorder labels See also Results section Study Population part note that only some of the ECGs used has a rhythm disorder see also “Performance of image-based multilabel classification” section second paragraph note that the training set only some of ECG had clinical labels). Re claim 3 Sangha discloses wherein the machine-learning based algorithm is a statistical algorithm (see “Performance of image-based multilabel classification“ 5th paragraph “For image data, we built a convolutional neural network (CNN) model based on the EfficientNet B3 architecture (see “Methods” for details)10. We trained our primary image model on both the standardized subset, with a 40%-40%-10%-10% split of standard, alternate, standard shuffled, and alternate shuffled images as well as the real-world variation subset” note that a convolution al neural network is used which can be considered a statistical algorithm). Re claim 4 Sangha discloses wherein the ECG image comprises a printed ECG image of an ECG dataset formed by conversion of ECG waveform data (see “Performance of image-based multilabel classification“ section “The ECGs were transformed to the corresponding printed images using the python library ecg-plot” note that ECG are generated by converting the waveform to printed image) Re claim 5 Sangha discloses wherein the method is generalizable to multiple ECG image formats (see discussion section first paragraph “We have developed an externally validated multilabel automated diagnosis algorithm that accurately identifies rhythm and conduction disorders from either ECG images or raw electrocardiographic signals. The algorithm demonstrates high discrimination and generalizes across two international waveform data sources, which are acquired on different equipment and temporally separated by over 2 decades”). Re claim 6 Sangha discloses wherein the algorithm trained on ECG images having incorrectly placed leads (see data processing section “The third and fourth formats were representations of the standard and alternate formats with shuffled lead locations. In the shuffled standard format precordial leads were presented in the first two columns and limb leads in the third and fourth. In shuffled alternative images precordial leads were presented in the first column and limb leads in the second (Fig. 1). This was done to prevent overfitting to certain locations on the ECG image and to create a more versatile algorithm” note that shuffling the leads corresponds to incorrectly placed leads). Re claim 7 Sangha discloses wherein the algorithm is trained on images of ECGs with different signal (See data processing last paragraph “First, each pre-image transformation sample started at a point that was selected from a uniform distribution of 0–1.3 s offset from the first detected QRS peak in lead I. This was designed to mimic the lack of predetermined starting points in real-world ECGs”, background (see data processing section last paragraph “The second variation introduced in the real-world variation dataset was different color schemes for the background of the ECG image and the signal being plotted. We modified the ecg-plot software to produce ECGs with three common color schemes to the ECGs in the ideal dataset. This included two variations of black and white ECGs, as well as a format with a pink background but a black signal (Fig. 1).”), and noise characteristics (see “Performance of image-based multilabel classification” last paragraph “we saw no drop off in the performance of the model as various types of noise were added”). Re claim 8 Sangha discloses comprising identifying hidden clinical labels (see discussion section first and 8th paragraph “we believe that our study identifies a strategy of leveraging ECG images for a broad set of disparate diagnoses—spanning rhythm and conduction disorders, as well as hidden labels”). Re claim 9 Sangha discloses further comprising identifying characteristics of the ECG image that the determination is based on (see discussion section 5th paragraph “Our examination of mean heatmaps across a sample of predictions for RBBB and LBBB demonstrates that classifications for these intraventricular conduction disturbances were guided by the same information human readers focus on when reading an ECG. Mean heatmaps consistently demonstrated the identification of specific leads that are important in the clinical diagnosis of RBBB and LBBB. Similar heatmaps applied to individual ECGs further supported similar interpretable learning across clinical labels. The identification of specific leads despite the variety of inputs in the training data, and the rotation of training images suggests that the model has learned more generalizable representations of ECG images, especially as it still identified clinically relevant leads in formats of images in the web-based real-world image dataset, which it had never encountered in model training.” note that heat maps are used to determine interpretable characteristics of the ECG image). Re claim 10 Sangha discloses wherein the method is automated. (see title). Re claim 11 Sangha discloses wherein the cardiovascular disease comprises a disorder selected from the group consisting of structural disorders of the heart, functional disorders of the heart, structural disorders of the structures supporting the heart, functional disorders of the structures supporting the heart, and combinations thereof (see discussion section first paragraph “We have developed an externally validated multilabel automated diagnosis algorithm that accurately identifies rhythm and conduction disorders from either ECG images or raw electrocardiographic signals” note that rhythm disorder would be a functional disorder of the heart). Re claim 12 Sangha discloses wherein the disorder comprises abnormalities of the muscle, valves, blood vessels, or lining of the heart. (see discussion section first paragraph “We have developed an externally validated multilabel automated diagnosis algorithm that accurately identifies rhythm and conduction disorders from either ECG images or raw electrocardiographic signals” note that rhythm disorder would be a functional disorder of the heart muscle). Re claim 16 Sangha discloses wherein, prior to the step of applying the algorithm to the ECG image for the subject (See Results section Study Population part note that only some of the ECGs used in the data sent has a rhythm disorder see also “Performance of image-based multilabel classification“ section second paragraph note that the training set only some of ECG had clinical labels), the method further comprises training the algorithm, the training of the algorithm comprising: creating an image-based dataset including a normal subset and a cardiovascular disease subset (see results first and second paragraph: “There were 2,228,236 ECGs from 1,506,112 patients acquired between 2010 and 2017 from 811 out of the 853 municipalities in the state of Minas Gerais, Brazil. The median age of the patients at the time of the ECG recording was 54 years (IQR 41, 67) and 60.3% of the ECGs were obtained among women. These ECGs were recorded as standard 12-lead recordings sampled at frequencies ranging from 300 to 600 Hz for 7– 10 s. In addition, information on patient demographics and 6 clinical labels were available (See “Methods” for details). Of these ECGs, 39,661 (1.8%) had a label for atrial fibrillation (AF), 61,551 (2.8%) for right bundle branch block (RBBB), 34,677 (1.6%) for left bundle branch block (LBBB), 34,446 (1.5%) for first-degree atrioventricular block (1dAVb), 35,441 (1.6%) for sinus bradycardia (SB), and 48,296 (2.2%) for sinus tachycardia (ST) (Supplementary Table 1). Of the 231,704 ECGs with at least one of the six detected rhythm disorders, 210,496 had exactly one, and 21,208 had more than one label.” Note that of the “2,228,236 ECGs” only “231,704 ECGs” had a label meaning some of the set had a disease label and some did not see also figure 4.); optionally pre-training the algorithm on an unrelated clinical or hidden label (see abstract “train a convolutional neural network (CNN) identifying 6 physician-defined clinical labels spanning rhythm and conduction disorders, and a hidden label for gender”); and training the algorithm on the image-based dataset ( see Performance of image-based multilabel classification sections “All ECGs were randomly subset into training, validation, and test sets (90%-5%-5%).” Also see abstract). Re claim 19 Sangha discloses wherein the clinical label includes six physician-defined labels and the hidden label includes gender (see abstract “A total of 2,228,236 12-lead ECGs signals from 811 municipalities in Brazil are transformed to ECG images in varying lead conformations to train a convolutional neural network (CNN) identifying 6 physician-defined clinical labels spanning rhythm and conduction disorders, and a hidden label for gender”) Re claim 22 Sangha discloses wherein the image-based dataset includes at least two different plotting schemes for each ECG waveform (See Performance of image-based multilabel classification Section “In the standardized subset, each pre-image transformation sample started 300 ms before the second QRS peak detected in lead I. To ensure that like a clinician the model learned about lead-specific information based on the label of the lead, rather than just the location of leads on a single ECG format, we included four formats of images in the standardized subset (Fig. 1). Standard printed ECGs in the United States typically consist of four 2.5 second columns presented sequentially on the page, with each column containing three leads of 2.5 s intervals of a continuous 10 s record. A 10-s rhythm strip was generated using a concatenation of lead I signal on the last and first QRS of the same 5 s signal7. The second plotting scheme we chose consisted of two columns each containing six 5-s recordings, with one column containing simultaneous limb leads and the other simultaneous precordial leads.” Note that multiple plotting schemes are used to plot the waveforms. Re claim 23 Sangha discloses wherein the image-based dataset includes at least two different ECG image formats (see Performance of image-based multilabel classification Section 3rd paragraph “To ensure that like a clinician the model learned about lead-specific information based on the label of the lead, rather than just the location of leads on a single ECG format, we included four formats of images in the standardized subset” note that the subset includes multiple formats). Claim Rejections - 35 USC § 103 The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. Claims 13 and 14 is/are rejected under 35 U.S.C. 103 as being obvious over Sangha et al “Automated multilabel diagnosis on electrocardiographic images and signals” Nature Communications volume 13, Article number: 1583 (2022) March 24 2022 in view of Albert US 2021/0121117. Re claim 13 Sangha discloses all the elements of claim 1. Sangha does not expressly disclose wherein the disorder is a genetic disorder. Albert discloses wherein the disorder is a genetic disorder (see abstract “Disclosed systems and method receive electrocardiogram (EKG) data of a subject. The EKG data comprises a record of at least a full beat of the subject. The EKG data is input into a machine learning model to generate an output. The output can include a segmentation of the full beat or a measurement of a QT interval of the subject” and see paragraph 4 “Long QT syndrome can be inherited genetically in some subjects. Genetically acquired long QT syndrome may be referred to as congenital long QT. While some of those subjects may not exhibit a long QT in all EKGs, they are still at higher risk for arrhythmias and other complications. Long QT can also be the result of environmental factors. Long QT resulting from environmental factors may be referred to as acquired long QT.” note that long QT can be and acquired or a genetic condition). The motivation to combine is “detection of and monitoring for long QT in individuals can be important when prescribing certain medications, changing exercise patterns, or changing other lifestyle patterns”. One of ordinary skill in the art could have easily modified the teachings of Sangha to also detect long QT syndrome. Therefore, it would have been obvious to one of orindary skill in the art before the effective filing date of the claimed invention to combine Alberta and Sangha to reach the aforementioned advantage. Re claim 14 Sangha discloses all the elements of claim 1. Sangha does not expressly disclose wherein the disorder is a acquired disorder. Albert discloses wherein the disorder is a acquired disorder (see abstract “Disclosed systems and method receive electrocardiogram (EKG) data of a subject. The EKG data comprises a record of at least a full beat of the subject. The EKG data is input into a machine learning model to generate an output. The output can include a segmentation of the full beat or a measurement of a QT interval of the subject” and see paragraph 4 “Long QT syndrome can be inherited genetically in some subjects. Genetically acquired long QT syndrome may be referred to as congenital long QT. While some of those subjects may not exhibit a long QT in all EKGs, they are still at higher risk for arrhythmias and other complications. Long QT can also be the result of environmental factors. Long QT resulting from environmental factors may be referred to as acquired long QT.” note that long QT can be and acquired or a genetic condition). The motivation to combine is “detection of and monitoring for long QT in individuals can be important when prescribing certain medications, changing exercise patterns, or changing other lifestyle patterns”. One of ordinary skill in the art could have easily modified the teachings of Sangha to also detect long QT syndrome. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Alberta and Sangha to reach the aforementioned advantage. Claim 15 17 and 18 is/are rejected under 35 U.S.C. 103 as being obvious over Sangha et al “Automated multilabel diagnosis on electrocardiographic images and signals” Nature Communications volume 13, Article number: 1583 (2022) March 24 2022 in view of Attia et al “Screening for cardiac contractile dysfunction using an artificial intelligence–enabled electrocardiogram” Nature Medicine volume 25, pages70–74 (2019). Re claim 15 claim 1 does not expressly disclose , wherein the cardiovascular disease comprises a disease that is not normally discernable by physicians from ECG data. Attia discloses (see main section “While strategies for early identification of ALVD may prevent progression to symptomatic heart failure, a noninvasive and low-cost screening tool does not currently exist” and “We hypothesized that the metabolic and structural derangements associated with the cardiomyopathic process would result in ECG changes that could be reliably detected by a properly trained neural network. To test this hypothesis, we created, trained, validated, and then tested a large neural network. The motivation to combine is “Asymptomatic left ventricular dysfunction (ALVD) is present in 3–6% of the general population, is associated with reduced quality of life and longevity, and is treatable when found1,2,3,4. An inexpensive, noninvasive screening tool for ALVD in the doctor’s office is not available. We tested the hypothesis that application of artificial intelligence (AI) to the electrocardiogram (ECG), a routine method of measuring the heart’s electrical activity, could identify ALVD” See abstract. One of ordinary skill could have easily modified the teachings of Sangra with the teachings of Attia to also detect ALVD. Therefore, it would have been obvious to one of orindary skill in the art before the effective filing date of the claimed invention to combine Attia and Sangha to reach the aforementioned advantage. Re claim 17 Sangha discloses all the elements of claim 16. Sangha does not disclose wherein the cardiovascular disease subset includes a low ejection fraction (EF) subset. Attia discloses wherein the cardiovascular disease subset includes a low ejection fraction (EF) subset (see abstract and “Main” section paragraph 3 note that a training set for EF disfunction is used). The motivation to combine is “Asymptomatic left ventricular dysfunction (ALVD) is present in 3–6% of the general population, is associated with reduced quality of life and longevity, and is treatable when found1,2,3,4. An inexpensive, noninvasive screening tool for ALVD in the doctor’s office is not available. We tested the hypothesis that application of artificial intelligence (AI) to the electrocardiogram (ECG), a routine method of measuring the heart’s electrical activity, could identify ALVD” See abstract. One of ordinary skill could have easily modified the teachings of Sangra with the teachings of Attia to also detect ALVD. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Attia and Sangha to reach the aforementioned advantage. Re claim 18 Attia further discloses wherein the low EF subset includes ECG images for individuals with EF of less than 40% (see abstract and “Main” section paragraph 3 note that a training set for EF disfunction is used with EF<35% which is also less that 40%) Claim 20 and 21 is/are rejected under 35 U.S.C. 103 as being obvious over Sangha et al “Automated multilabel diagnosis on electrocardiographic images and signals” Nature Communications volume 13, Article number: 1583 (2022) March 24 2022 in view of Bhattacharya et al “Machine Learning Methods for Identifying Atrial Fibrillation Cases and Their Predictors in Patients With Hypertrophic Cardiomyopathy: The HCM-AF-Risk Model” CJC 2021. Re claim 20 Sangha discloses all the elements of claim 16. Sangha does not disclose wherein the normal subset includes ECG images for individuals having hypertrophic cardiomyopathy (HCM). Bhattacharya discloses wherein the normal subset includes ECG images for individuals having hypertrophic cardiomyopathy (HCM) see “Figure 2. Methods for addressing data imbalance—the illustration shows our classification scheme for combining oversampling and undersampling. The topmost layer represents the entire training set, which comprises a majority of No-atrial fibrillation (AF) records (left) and the minority of AF records (right). The majority class in the training set (No-AF) is randomly undersampled such that the No-AF to AF record ratio is 2:1. The minority class (AF) is oversampled using Synthetic Minority Oversampling Technique (SMOTE) to generate synthetic new AF-like records, doubling the number of AF records. The resulting set forms a balanced training set, containing the same number of AF and No-AF records.” Note that data sets include HCM with no AF and HCM with AF. The motivation to combine is “Hypertrophic cardiomyopathy (HCM) patients have a high incidence of atrial fibrillation (AF) and increased stroke risk, even with low CHA2DS2-VASc (congestive heart failure, hypertension, age diabetes, previous stroke/transient ischemic attack) scores. Hence, there is a need to understand the pathophysiology of AF/stroke in HCM. In this retrospective study, we develop and apply a data-driven, machine learning–based method to identify AF cases, and clinical/imaging features associated with AF, using electronic health record data.” (See background). One of ordinary skill in the art could have easily modified Sangha with Bhattacharya to additionally check for atrial fibrillation in HCM patients. Therefore it would have been obvious to one of orindary skill in the art before the effective filing date of the claimed invention to combine Bhattacharya and Sangha to reach the aforementioned advantage. Re claim 21 Sangha does not disclose wherein the cardiovascular disease subset includes ECG images for individuals having HCM and left ventricular (LV) systolic dysfunction. Bhattacharya discloses wherein the cardiovascular disease subset includes ECG images for individuals having HCM and left ventricular (LV) systolic dysfunction. “the HCM-AF-Risk Model is the first machine learning–based method for the identification of AF cases and clinical/imaging features associated with higher/lower risk of AF in HCM, using electronic health record data. In our model, individual patient data are represented as an N-dimensional vector, and the model output is a probability score for AF (AF risk) in HCM. We identified 18 clinical variables that are highly associated (positively/negatively) with AF in HCM patients. In addition to age, NYHA class, LA size, and LV fibrosis, which have been previously found to be associated with AF in HCM,29, 30, 31 we found that additional clinical features, such as LV diastolic dysfunction and lower LV-systolic strain, are positively associated with AF, and greater exercise capacity is negatively associated with AF in HCM”. The motivation to combine is “Hypertrophic cardiomyopathy (HCM) patients have a high incidence of atrial fibrillation (AF) and increased stroke risk, even with low CHA2DS2-VASc (congestive heart failure, hypertension, age diabetes, previous stroke/transient ischemic attack) scores. Hence, there is a need to understand the pathophysiology of AF/stroke in HCM. In this retrospective study, we develop and apply a data-driven, machine learning–based method to identify AF cases, and clinical/imaging features associated with AF, using electronic health record data.” (See background). One of ordinary skill in the art could have easily modified Sangha with Bhattacharya to additionally check for atrial fibrillation in HCM patients. Therefore, it would have been obvious to one of orindary skill in the art before the effective filing date of the claimed invention to combine Bhattacharya and Sangha to reach the aforementioned advantage. Cited Art The following is a listing of cited art considered relevant but not cited in a rejection above: WAGNER US 20220189636 A1 discloses isclosed systems, methods, and computer readable media can diagnose a health condition based on patient time series data. For example, a method for diagnosing a health condition based on patient time series data includes identifying a training set of health records comprising a first set of patient time series data, training a neural network using the training set of health records, and executing the trained neural network model to diagnose a health condition based on a second set of patient time series data. In further examples, the first set of patient time series data and the second set of patient time series data can each comprise electrocardiogram data and the health condition can comprise pulmonary hypertension. (see abstract)’ Bruce US 20230253114 A1 discloses A method for controlling operations of a computer system for providing medical diagnoses, comprises diagnosing a patient, wherein the diagnosing comprises receiving a patient identification of the patient; receiving electronic data indicating the geographic location of the computer system; receiving the patient's chief complaint and interview data; receiving electronic data indicating one or more current body characteristics of the patient comprising at least one of pulse rate, body temperature, blood pressure, respiration, and skin condition; creating a current multimedia representation for each of the one or more current body characteristics determined by using the one or more sensors; identifying, in the computer system, previous multimedia representations of each of the one or more body characteristics from other persons using one or more trained classifiers; comparing the current multimedia representation to the previous multimedia representations; and identifying potential matches with corresponding confidence factors in accordance with defined medical standards. (see abstract). Johnson US 20220183571 A1 discloses Computer-implemented systems and methods are provided for supplying electrocardiograms and identified patient information to an artificial intelligence engine comprising a neural network configured with a fractional flow reserve prediction model and that predicts a calculated fractional flow reserve for the patient, from which a predicted occurrence of one or more cardiac events is determined (see abstract). Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to SEAN T MOTSINGER whose telephone number is (571)270-1237. The examiner can normally be reached 9AM-5PM. 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, Chineyere Wills-Burns can be reached at (571) 272-9752. 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. /SEAN T MOTSINGER/Primary Examiner, Art Unit 2673
Read full office action

Prosecution Timeline

Nov 27, 2024
Application Filed
Nov 04, 2026
Non-Final Rejection mailed — §101, §102, §103 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12802278
Fisheye image compression method, fisheye video stream compression method and panoramic video generation method
2y 10m to grant Granted Nov 17, 2026
Patent 12784423
CAMERA POSE RELATIVE TO OVERHEAD IMAGE
2y 7m to grant Granted Nov 03, 2026
Patent 12784209
SYSTEM AND METHOD FOR CONVERTING SKIN TISSUE IMAGES BASED ON DEEP LEARNING
2y 4m to grant Granted Nov 03, 2026
Patent 12776044
IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, AND IMAGE PROCESSING PROGRAM
1y 11m to grant Granted Oct 27, 2026
Patent 12770382
METHOD AND SYSTEM FOR SCREENING HIGH DEVELOPMENTAL POTENTIAL EMBRYO FOR IN-VITRO FERTILIZATION
2y 2m to grant Granted Oct 20, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

Strategy Recommendation AI-generated — please review before filing

Get a prosecution strategy drawn from examiner precedents, rejection analysis, and claim mapping.
Typically takes 5-10 seconds — AI-generated, attorney review required before filing

Prosecution Projections

1-2
Expected OA Rounds
79%
Grant Probability
90%
With Interview (+11.8%)
2y 11m (~11m remaining)
Median Time to Grant
Low
PTA Risk
Based on 702 resolved cases by this examiner. Grant probability derived from career allowance rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month