DETAILED ACTION
1. This communication is in response to the amendments filed on September 16, 2025 for Application No. 18/878,007 in which Claims 1, 5-11, 13-16, and 61 are presented for examination.
Notice of Pre-AIA or AIA Status
2. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
3. The amendments filed on September 16, 2025 have been considered. Claims 1 and 7 have been amended. Claim 4 has been cancelled. Claim 61 has been newly added. Thus, Claims 1, 5-11, 13-16, and 61 are pending and presented for examination.
4. Applicant’s arguments and amendments filed September 16, 2025 with respect to the 35 U.S.C. 112(b) rejection have been fully considered and are persuasive. The 35 U.S.C. 112(b) rejection has been withdrawn.
5. Applicant's arguments filed September 16, 2025 with respect to the 35 U.S.C. 101 rejection have been fully considered but they are not persuasive.
Applicant’s Arguments on Pg. 6 of Arguments/Remarks state:
“The Examiner rejected claims 1, 5-11, and 13-16 under 35 U.S.C. §101 as being directed to an abstract idea. Although applicant disagrees, applicant has amended claim 1 to address the Examiner's concern.
Claim 1 now recites that the machine learning (ML) model is employed to identify a patent derived region given a patient base region that is manually selected. The patient is then treated based on the patient derived region.
These limitations integrate the alleged abstract idea into a specific, practical application - namely, a medical treatment method that results in a tangible outcome for a patient. Such a claim falls squarely within the categories of patent-eligible subject matter as outlined in the Mayo/Alice framework Step 2A, Prong Two, and MPEP §2106.
Because amended claim 1 is directed to a specific medical treatment process and does not preempt the abstract idea itself, applicant respectfully submits that the claims are patent-eligible and that the §101 rejection should be withdrawn.”
Examiner respectfully disagrees. The disclosed machine learning model, and its corresponding training, is still recited at a high-level of generality and does not integrate the abstract idea into practical application. More specifically, Claim 1 recites “training the ML model using the training data wherein the trained ML model, when applied to a subject region of a subject cardiogram, outputs base data as subject data for the subjection region” and “wherein a patient base region of a patient cardiogram collected from a patient is input to the ML model which outputs a patient derived region and the patient is treated based on the patient derived cardiogram” without significantly more. First, the training of the ML model is merely applied, as the claim simply recites that the model is “trained” simply “using training data”, followed by subsequent language that merely recites how the model functions once trained without actually detailing how the model is trained to perform such specific operations just by “using training data” in the first place. This applies similarly to the newly added limitation – the trained ML model accepts an input and further outputs a “patient derived region” which is somehow used to treat the patient without significantly more. These inputting/outputting steps are well-understood, routine, conventional activity at Step 2A Prong 2 and Step 2B and further, specifying that the patient is treated based on the “patient derived cardiogram” amounts to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. The remaining limitations recite a plurality of mental process steps at Step 2A Prong 1, as well as additional insignificant extra-solution activities and merely indicating field of use at Step 2A Prong 2 and Step 2B. The updated 35 U.S.C. 101 rejection may be found in the subsequent section below.
Thus, the 35 U.S.C. 101 rejection is maintained.
6. Applicant’s arguments filed September 16, 2025 with respect to the 35 U.S.C. 102 rejection have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. Examiner has updated the 35 U.S.C. 102 rejection to a 35 U.S.C. 103 rejection unpatentable over Anastasia in view of Baram, where Baram teaches the amended/argued claim limitations. The updated rejection may be found in the subsequent 35 U.S.C. 103 section below.
Specification
7. Applicant is reminded of the proper content of an abstract of the disclosure.
A patent abstract is a concise statement of the technical disclosure of the patent and should include that which is new in the art to which the invention pertains. The abstract should not refer to purported merits or speculative applications of the invention and should not compare the invention with the prior art.
If the patent is of a basic nature, the entire technical disclosure may be new in the art, and the abstract should be directed to the entire disclosure. If the patent is in the nature of an improvement in an old apparatus, process, product, or composition, the abstract should include the technical disclosure of the improvement. The abstract should also mention by way of example any preferred modifications or alternatives.
Where applicable, the abstract should include the following: (1) if a machine or apparatus, its organization and operation; (2) if an article, its method of making; (3) if a chemical compound, its identity and use; (4) if a mixture, its ingredients; (5) if a process, the steps.
Extensive mechanical and design details of an apparatus should not be included in the abstract. The abstract should be in narrative form and generally limited to a single paragraph within the range of 50 to 150 words in length.
See MPEP § 608.01(b) for guidelines for the preparation of patent abstracts.
Claim Rejections - 35 USC § 101
8. 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.
9. Claims 1, 5-11, 13-16, and 61 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding Claim 1:
Step 1: Claim 1 is a method type claim. Therefore, Claims 1, 5-11, 13-16, and 61 are directed to either a process, machine, manufacture, or composition of matter.
2A Prong 1: 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. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation by mathematical calculation but for the recitation of generic computer components, then it falls within the “Mathematical Concepts” grouping of abstract ideas.
[…] identifying data associated with a region of a cardiogram […] (mental process – identifying data associated with a region of a cardiogram may be performed manually by a user observing/analyzing the cardiogram and accordingly using judgement/evaluation to identify data associated with a particular region of said cardiogram)
for each of the plurality of mappings, deriving a plurality of derived regions from the base region and the base cardiogram of that mapping (mental process – deriving a plurality of regions from the base region and base cardiogram may be performed manually by a user observing/analyzing the plurality of mappings that each map a base region of a base cardiogram to base data and accordingly using judgement/evaluation to derive a plurality of derived regions based on the user’s review of each of the mappings)
for each of the plurality of the derived regions, generating training data that includes the derived region labeled with the base data of that mapping (mental process – generating training data may be performed manually by a user observing/analyzing each of the derived regions and accordingly using judgement/evaluation to generate training data which includes the labeled derived region with the base data)
[…] adding a positive or negative increment to the base start time of that base region and/or adding a positive or negative increment to the base end time of that base region (mental process – adding a positive or negative increment to the base start/end time may be performed manually by a user using judgment/evaluation to either add a positive or negative increment to either the base start time and/or base end time)
[…] the patient base region being selected by a person (mental process – selecting a patient base region may be performed manually by a user observing/analyzing the cardiogram and accordingly using judgement/evaluation to select a base region of the cardiogram)
2A Prong 2: This judicial exception is not integrated into a practical application.
Additional elements:
one or more computing systems for training a machine learning model (ML) (recited at a high-level of generality (i.e., as a generic computing system for generically training a machine learning model without significantly more) such that it amounts to no more than mere instructions to apply the exception using generic computer components)
accessing a plurality of mappings that each map a base region of a base cardiogram to base data (Adding insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g))
training the ML model using the training data (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training a machine learning model with previously determined data)
wherein the trained ML model, when applied to a subject region of a subject cardiogram, outputs base data as subject data for the subject region (Adding insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g))
wherein a base region of a base cardiogram has a base start time and a base end time within that base cardiogram and wherein the derived regions that are derived from a base region have derived start times and derived end times that are derived from the base start time and the base end time of that base region (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the base region of the cardiogram has a base start/end time and the derived regions have a derived start/end time does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
wherein the derived start times and the derived end times for derived regions that are derived from a base region are derived by adding a positive or negative increment to the base start time of that base region and/or adding a positive or negative increment to the base end time of that base region (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the derived start/end times are derived from a base region by adding a positive or negative increment to the base start/end time does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
wherein a patient base region of a patient cardiogram collected from a patient is input to the ML model […] (Adding insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g))
[…] which outputs a patient derived region […] (Adding insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g))
[…] the patient is treated based on the patient derived cardiogram (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the patient is treated simply based on the patient derived cardiogram does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
one or more computing systems for training a machine learning model (ML) (mere instructions to apply the exception using generic computer components cannot provide an inventive concept)
accessing a plurality of mappings that each map a base region of a base cardiogram to base data (MPEP 2106.05(d)(II) indicates that merely “Receiving or transmitting data over a network” is a well-understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed limitation is well-understood, routine, conventional activity is supported under Berkheimer)
training the ML model using the training data (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training a machine learning model with previously determined data)
wherein the trained ML model, when applied to a subject region of a subject cardiogram, outputs base data as subject data for the subject region (MPEP 2106.05(d)(II) indicates that merely “Presenting offers and gathering statistics” is a well-understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed limitation is well-understood, routine, conventional activity is supported under Berkheimer)
wherein a base region of a base cardiogram has a base start time and a base end time within that base cardiogram and wherein the derived regions that are derived from a base region have derived start times and derived end times that are derived from the base start time and the base end time of that base region (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the base region of the cardiogram has a base start/end time and the derived regions have a derived start/end time does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
wherein the derived start times and the derived end times for derived regions that are derived from a base region are derived by adding a positive or negative increment to the base start time of that base region and/or adding a positive or negative increment to the base end time of that base region (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the derived start/end times are derived from a base region by adding a positive or negative increment to the base start/end time does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
wherein a patient base region of a patient cardiogram collected from a patient is input to the ML model […] (MPEP 2106.05(d)(II) indicates that merely “Receiving or transmitting data over a network” is a well-understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed limitation is well-understood, routine, conventional activity is supported under Berkheimer)
[…] which outputs a patient derived region […] (MPEP 2106.05(d)(II) indicates that merely “Presenting offers and gathering statistics” is a well-understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed limitation is well-understood, routine, conventional activity is supported under Berkheimer)
[…] the patient is treated based on the patient derived cardiogram (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the patient is treated simply based on the patient derived cardiogram does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
For the reasons above, Claim 1 is rejected as being directed to an abstract idea without significantly more. This rejection applies equally to dependent claims 5-11, 13-16, and 61. The additional limitations of the dependent claims are addressed below.
Regarding Claim 5:
Step 2A Prong 1:
See the rejection of Claim 1 above, which Claim 5 depends on.
Step 2A Prong 2 & Step 2B:
wherein the ML model comprises a convolutional neural network that inputs an image of a derived region and outputs the base data (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the ML model is a convolutional neural network that inputs an image and outputs base data does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 6:
Step 2A Prong 1: See the rejection of Claim 1 above, which Claim 6 depends on.
generates a latent vector representing an image of the derived region […] (mathematical process – generating a latent vector representing an image of the derived region may be performed by mathematical process/calculation using an equation/algorithm for generating a latent vector)
Step 2A Prong 2 & Step 2B:
wherein the ML model comprises a portion of autoencoder […] and another ML model that inputs the generated latent vector and outputs the base data (mere instructions to apply the exception using generic computer components cannot provide an inventive concept)
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 7:
Step 2A Prong 1:
See the rejection of Claim 1 above, which Claim 7 depends on.
Step 2A Prong 2 & Step 2B:
wherein the ML model is trained using features derived from the derived region (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner' s note: high level recitation of training a machine learning model with previously determined data)
the features being one or more of a time-voltage series specifying voltages and time increments of the derived region, images and time- voltage series of portions of the derived region, length in seconds of various intervals of the derived region, QRS integral of the derived region, maximum, minimum, mean, and variance of voltages of the derived region, a maximal vector of QRS loop and angle of a vector derived from vectorcardiogram of the derived region, and location of a peak or zero crossing relative to a maximum peak in an interval of the derived region (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying the features without significantly more does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 8:
Step 2A Prong 1: See the rejection of Claim 1 above, which Claim 8 depends on.
[…] generate an encoding […] (mathematical process – generating an encoding may be performed by mathematical process using an algorithm/equation for generating an encoding)
Step 2A Prong 2 & Step 2B:
wherein the ML model comprises an encoder of a transformer adapted to an image to […] and a neural network inputs the encoding and other features of a feature vector based on the training data (mere instructions to apply the exception using generic computer components cannot provide an inventive concept)
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 9:
Step 2A Prong 1:
See the rejection of Claim 1 above, which Claim 9 depends on.
Step 2A Prong 2 & Step 2B:
wherein the ML model comprises a neural network that inputs a feature vector representing the derived region (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the ML model comprises a neural network that inputs a feature vector does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 10:
Step 2A Prong 1:
See the rejection of Claim 9 above, which Claim 10 depends on.
Step 2A Prong 2 & Step 2B:
wherein the feature vector representing the derived region includes a time-voltage series of the derived region (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the feature vector represents the derived region including a time-voltage series does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 11:
Step 2A Prong 1:
See the rejection of Claim 1 above, which Claim 11 depends on.
Step 2A Prong 2 & Step 2B:
wherein at least some of the mappings have a base time range of the base cardiogram that is selected by a person and have base data that is specified based on treatment of an arrhythmia (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the mappings have a base time range selected by a person and have base data that is specified base on treatment of an arrhythmia does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 13:
Step 2A Prong 1: See the rejection of Claim 1 above, which Claim 13 depends on.
determining overall subject data based on analysis of the lead subject data for the leads (mental process – determining overall subject data may be performed manually by a user observing/analyzing the lead subject data for the leads and accordingly using judgement/evaluation to determine overall subject data)
Step 2A Prong 2 & Step 2B:
wherein a base cardiogram includes multiple leads and wherein a base region of each lead is mapped to base data and wherein a machine learning model is trained for each lead (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the base cardiogram includes multiple leads and wherein a base region of each lead is mapped to base data and wherein a machine learning model is trained for each lead does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
receiving lead subject regions of multiple leads of a subject cardiogram (MPEP 2106.05(d)(II) indicates that merely “Receiving or transmitting data over a network” is a well-understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed limitation is well-understood, routine, conventional activity is supported under Berkheimer)
for each of the multiple leads, applying a trained machine learning model for that lead to the lead subject region of that lead, which outputs lead subject data for that lead(Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner' s note: high level recitation of training a machine learning model with previously determined data)
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 14:
Step 2A Prong 1:
See the rejection of Claim 1 above, which Claim 14 depends on.
Step 2A Prong 2 & Step 2B:
wherein the base data is a source location of an arrhythmia (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the base data is a source location of an arrhythmia does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 15:
Step 2A Prong 1:
See the rejection of Claim 1 above, which Claim 15 depends on.
Step 2A Prong 2 & Step 2B:
wherein the base data is a region of a cardiogram (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the base data is a region of a cardiogram does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 16:
Step 2A Prong 1:
See the rejection of Claim 1 above, which Claim 16 depends on.
Step 2A Prong 2 & Step 2B:
wherein the base data is a start time and an end time of a region a cardiogram (Field of Use – limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception does not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application; in this case specifying that the base data is a start time and an end time of a region of a cardiogram does not integrate the exception into a practical application nor amount to significantly more – See MPEP 2106.05(h))
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Regarding Claim 61:
Step 2A Prong 1: See the rejection of Claim 1 above, which Claim 61 depends on.
wherein the treating of the patient includes identifying a source location of an arrhythmia based on the patient derived cardiogram and performing an ablation on the patient based on the identified source location (mental process – identifying a source location of an arrhythmia and performing an ablation may be performed manually by a user observing/analyzing the patient derived cardiogram and using judgement/evaluation to identify a source location of an arrhythmia and accordingly performing an ablation based on the located arrhythmia)
Step 2A Prong 2 & Step 2B:
Accordingly, under Step 2A Prong 2 and Step 2B, these additional elements do not integrate the abstract idea into practical application because they do not impose any meaningful limits on practicing the abstract idea, as discussed above in the rejection of claim 1.
Claim Rejections - 35 USC § 103
10. 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.
11. Claims 1, 5-11, 13-16, and 61 are rejected under 35 U.S.C. 103 as being unpatentable over Anastasia et al. (hereinafter Anastasia) (US PG-PUB 20210315506), in view of Baram et al. (hereinafter Baram) (US PG-PUB 20220000410).
Regarding Claim 1, Anastasia teaches a method performed by one or more computing systems for training a machine learning model (ML) for identifying data associated with a region of a cardiogram (Anastasia, Par. [0004], “In accordance with various embodiments, there is provided a computer-implemented method of facilitating electrocardiogram (“ECG”) analysis. […] The method involves causing at least one neural network classifier to be applied to the one or more determined representative ECG traces to determine one or more diagnostically relevant scores related to at least one diagnosis of the patient.”, therefore, methods for training a neural network/machine learning model for identifying data associated with a region of an ECG are disclosed), the method comprising:
accessing a plurality of mappings that each map a base region of a base cardiogram to base data (Anastasia, Claim 14, “receiving a plurality of sets of training ECG traces, wherein each set of the sets of training ECG traces represents sensed heart activity over a training time period for a respective associated training patient of a plurality of training patients; receiving, for each set of the plurality of sets of training ECG traces, a respective diagnosis for the training patient associated with the set of training ECG traces;”, therefore, a plurality of sets of training ECG traces which map a base region (represented by the ECG traces) of an ECG to base data (a respective diagnosis for the associated ECG trace) are accessed)
for each of the plurality of mappings (Anastasia, Claim 14, “for each of the training ECG traces: […]”, therefore, the subsequent operations are performed for each of the plurality of mappings/training ECG traces),
deriving a plurality of derived regions from the base region and the base cardiogram of that mapping (Anastasia, Claim 14, “identifying a plurality of corresponding training ECG trace segments, each of the training ECG trace segments representing patient heart activity over a segment of the training time period; and”, therefore, a plurality of regions/segments over a segment of a period of time are derived (ECG trace segments) from the complete base region/ECG trace); and
for each of the plurality of the derived regions, generating training data that includes the derived region labeled with the base data of that mapping (Anastasia, Claim 14, “determining a representative training ECG trace based on at least one of the identified corresponding training ECG trace segments; and”, thus, for each of the plurality of derived regions/segments, training data including a training ECG trace based on at least one of the derived segments/regions is generated); and
training the ML model using the training data wherein the trained ML model, when applied to a subject region of a subject cardiogram, outputs base data as subject data for the subject region (Anastasia, Claim 14, “causing at least one neural network classifier to be trained using the representative training ECG traces and the diagnoses, the at least one neural network classifier configured to output one or more diagnostically relevant scores related to at least one diagnosis:,”, therefore, a machine learning model/neural network classifier is trained using the training ECG traces and diagnoses, such that when applied to a subject region of a cardiogram, the classifier outputs base data (diagnosis data) for the subject region)
wherein a base region of a base cardiogram has a base start time and a base end time within that base cardiogram and wherein the derived regions that are derived from a base region have derived start times and derived end times that are derived from the base start time and the base end time of that base region (Anastasia, Claim 22, “The method of claim 14 wherein identifying the plurality of corresponding training ECG trace segments comprises identifying respective common features in the training ECG trace segments and identifying respective start and end times for each of the plurality of training ECG trace segments relative to the identified common features.”, thus, each ECG trace segment, including both the base ECG traces and segmented/derived ECG traces, has an associated start time and end time),
wherein the derived start times and the derived end times for derived regions that are derived from a base region are derived by adding a positive or negative increment to the base start time of that base region and/or adding a positive or negative increment to the base end time of that base region (Anastasia, Par. [0106], “In various embodiments, block 304 may direct the analyzer processor 100 to identify the start and end times by subtracting and adding, respectively, half of the temporal length of the window from each R-peak identifier.”, therefore, the segmented/derived start and end times may be identified by subtracting and adding to the base times), and
wherein a patient base region of a patient cardiogram collected from a patient is input to the ML model which outputs a patient derived region and the patient is treated based on the patient derived cardiogram (See introduction of Baram reference below for teaching of inputting a patient cardiogram to an ML model which outputs a derived region), the patient base region being selected by a person (Anastasia, Par. [0129], “ For example, in some embodiments, the ECG trace from which the ECG segments are derived may be a very short trace, including only a few complete heartbeats and, in these cases the statistical measures (PCA for example) may become less reliable due to a limited amount of data. In such cases, a single heartbeat may be selected arbitrarily as the representative ECG trace, for example based on the operator's judgment, though this may be avoided if at all possible.”, therefore, the patient base region/ECG trace may be selected by an operator).
While Anastasia discloses training an ML model, wherein the trained ML model outputs base data as subject data for the subject region as shown by the claim mapping above, Anastasia does not explicitly disclose wherein a patient base region of a patient cardiogram collected from a patient is input to the ML model which outputs a patient derived region and the patient is treated based on the patient derived cardiogram.
However, Baram teaches wherein a patient base region of a patient cardiogram collected from a patient is input to the ML model (Baram, Claim 10, “[…] a neural network that operates on the spatial ECG feature vectors and the spatial shape representations to produce a plurality of outputs.”, thus, a neural network takes in spatial electrocardiogram features representing a base region of a patient cardiogram as input. This is similarly supported by Par. [0170] which mentions that recorded ECG signals may be provided as inputs to the system) which outputs a patient derived region (Baram, Claim 13, “The system of claim 10 wherein the outputs include a next region of interest.”, therefore, the neural network may output a patient derived region/region of interest. This is similarly supported by Par. [0170] which mentions that the system may output a next region of interest), and the patient is treated based on the patient derived cardiogram (Baram, Par. [0040], “Examples of patient biometrics include electrical signals (e.g., ECG signals and brain biometrics), blood pressure data, blood glucose data and temperature data. The patient biometrics may be monitored and communicated for treatment across any number of various diseases, such as cardiovascular diseases (e.g., arrhythmias, cardiomyopathy, and coronary artery disease) and autoimmune diseases (e.g., type I and type II diabetes).”, thus, the patient may be treated based on the region of interest of the ECG/derived cardiogram).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the method for training a machine learning model for identifying data associated with a region of a cardiogram, as disclosed by Anastasia to include wherein a patient base region of a patient cardiogram collected from a patient is input to the ML model which outputs a patient derived region and the patient is treated based on the patient derived cardiogram, as disclosed by Baram. One of ordinary skill in the art would have been motivated to make this modification to enable the machine learning model to learn from the most informative derived region, hence improving system efficiency and accuracy (Baram, Par. [0030], “Although the details of the application will be described herein, briefly machine learning (ML) is utilized to train a system to improve mapping efficiency by guiding the user to acquire points in informative locations.” & Par. [0167], “A confidence level 2090 measured by locally consistence of signal representations and amount of meaningful data. By sampling points over the anatomy surface using the network, outputs may include point(s) in space indicating the next most informative region to sample. These point(s) may be from a region having high confidence values 2090 and a point in space representing approximated new origin using intersection of vector to source or using the classical triangulation algorithm.”).
Regarding Claim 5, Anastasia in view of Baram teaches the method of claim 1, wherein the ML model comprises a [[convolutional]] neural network that inputs an image of a derived region and outputs the base data (Anastasia, Par. [0004], “The method involves receiving one or more sensed ECG traces for a patient, each of the sensed ECG traces representing sensed patient heart activity over a sensed time period, and, for each of the one or more sensed ECG traces: identifying a plurality of corresponding sensed ECG trace segments, each of the sensed ECG trace segments representing sensed patient heart activity for the patient over a segment of the sensed time period, and determining a representative ECG trace based on at least one of the identified corresponding sensed ECG trace segments. The method involves causing at least one neural network classifier to be applied to the one or more determined representative ECG traces to determine one or more diagnostically relevant scores related to at least one diagnosis of the patient.”, therefore, the machine learning model comprising a neural network may include inputting of an image (graphical ECG trace) and outputting of base data (patient diagnosis data)).
Anastasia does not explicitly disclose the neural network being a convolutional neural network.
However, Baram teaches a convolutional neural network (Baram, Par. [0162], “A second NN 2030 may operate on the anatomy surface information 2010.3 and other inputs 2010. The anatomy surface may be represented using a Vnet network This network may be a fully convolutional network with residual connections.”, thus, a convolutional neural network is disclosed).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the method of claim 1 for training a machine learning model for identifying data associated with a region of a cardiogram, as disclosed by Anastasia in view of Baram to include the use of a convolutional neural network, as disclosed by Baram. One of ordinary skill in the art would have been motivated to make this modification to enable improved processing of images, including more efficient identification of data within a cardiogram, as convolutional neural networks better depict spatial locations and relationships as compared to other types of neural networks (Baram, Par. [0162], “The anatomy surface may be represented using a Vnet network This network may be a fully convolutional network with residual connections. This network accepts an input volume of a predefined size (50×50×50 for our case) and returns an output volume of the same size. The first part performs 3D convolutions followed by a non-linearity (RELU) and down-sampling (Max-pooling) in each layer. Each 3D convolution has filters of size 50×50×50×C, where C is the number of channels. The weights of the first three dimensions represent the spatial location and are shared, while C is the current dimension of input or feature data. The inner part (vector of size 512 for the network in the figure) represents the latent space with the addition (via concatenation) of every residual link.”).
Regarding Claim 6, Anastasia in view of Baram teaches the method of claim 1.
Anastasia does not explicitly disclose wherein the ML model comprises a portion of autoencoder that generates a latent vector representing an image of the derived region and another ML model that inputs the generated latent vector and outputs the base data.
However, Baram teaches wherein the ML model comprises a portion of autoencoder that generates a latent vector representing an image of the derived region (Baram, Par. [0158], “The recorded ECG signals and location 2010.1 and Carto annotations may be combined in a first NN 2020. This input may also include other inputs 2010. A transformer network 2020 may be used to encode the ECG signals into a feature vector. […] The transformer itself is built from an encoder and decoder, thus defining the latent vector as the input to the encoder.”, therefore, the machine learning model comprises a portion of autoencoder that generates a latent vector representing an image) and another ML model that inputs the generated latent vector and outputs the base data (Baram, Par. [0162], “A second NN 2030 may operate on the anatomy surface information 2010.3 and other inputs 2010. The anatomy surface may be represented using a Vnet network This network may be a fully convolutional network with residual connections.”, thus, another machine learning model that inputs the latent vector and outputs base data is disclosed).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the method of claim 1 for training a machine learning model for identifying data associated with a region of a cardiogram, as disclosed by Anastasia in view of Baram to include wherein the ML model comprises a portion of autoencoder that generates a latent vector representing an image of the derived region and another ML model that inputs the generated latent vector and outputs the base data, as disclosed by Baram. One of ordinary skill in the art would have been motivated to make this modification to enable the representation of information in a latent space by a latent vector, which may capture important features, hence improving model performance (Baram, Par. [0157], “For the system 2000 to be useful for an intra-patient anatomical and activation propagation variety, the system 2000 may learn to extract uniform, compact, and meaningful information about the anatomy and ECG signals. According to an embodiment, two networks that can perform this operation in an unsupervised manner. The information may be represented in a latent space Rk (where K is fixed and relatively small) which maps the input from its original large vector dimension into a more compact one and reconstructs the input using only the latent representation. The latent space can capture important features by itself or tuned by additional tasks and constrains”).
Regarding Claim 7, Anastasia in view of Baram teaches the method of claim 1 wherein the ML model is trained using features derived from the derived region the features being one or more of a time-voltage series specifying voltages and time increments of the derived region (Anastasia, Par. [0012], “Identifying the plurality of corresponding sensed ECG trace segments may involve identifying respective common features in the sensed ECG trace segments and identifying respective start and end times for each of the plurality of sensed ECG trace segments relative to the identified common features.”, therefore, the machine learning model may be trained utilizing common features derived from the corresponding derived/segmented ECG traces. This may include time increments/start and end times relative to the common features), images and time- voltage series of portions of the derived region, length in seconds of various intervals of the derived region, QRS integral of the derived region, maximum, minimum, mean, and variance of voltages of the derived region, a maximal vector of QRS loop and angle of a vector derived from vectorcardiogram of the derived region, and location of a peak or zero crossing relative to a maximum peak in an interval of the derived region (Anastasia, Par. [0105], “Block 304 then directs the analyzer processor 100 to identify respective start and end times for each of the plurality of sense