DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 07/28/2025 has been entered.
Response to Arguments
Applicant’s arguments, see page 1, filed 07/28/2025, with respect to 35 U.S.C. 112(b) have been fully considered and are persuasive. The rejection of the claims has been withdrawn.
Applicant's arguments, see pages 1-10, filed 07/28/2025, regarding 35 U.S.C. 101 have been fully considered but they are not persuasive. Applicant asserts on page 3 that the claims cannot be practically performed in the human mind and argues on page 5 that “These steps involve ingesting and processing high-volume, multi-channel physiological data with temporal dependencies that cannot be mentally performed. Moreover, the classification of each time series segment to one or more labels using a trained time series classified, based on training data comprising correlated examples of segment attributes and labels, further underscores the technical nature of the claimed process. These steps necessitate machine learning algorithms, statistical modeling, and signal pattern recognition tools that, as described in the specification, require computing resources and specialized models. Finally, the generation of a time series report comprising a chronological analysis of IEGM segments and correlation to clinical cardiac anomalies involves organizing and interpreting labeled outputs to reveal diagnostic event patterns, another process clearly beyond the capacity of the human mind alone. Consistent with the guidance in Jn re Cortright, In re Suitco, and In re Buszard, one of ordinary skill in the art of biomedical signal processing would not interpret these limitations as encompassing mere mental steps.” Examiner disagrees because the limitations cites “employs a temporal alignment technique to synchronize the segments across multiple channels of data…identify one or more segment attributes for each time series segment of the plurality of time series segments… classifying each time series segment of the plurality of time series segments to at least one time series label using the trained time series classifier” which can all be done by a human mind because determining an alignment can be observed by a human, identifying features can be done by observation, and classifying by labeling trained data can be done via evaluation. The use of a processor and algorithms are simply automating the abstract idea on a computer.
Applicant then argues on page 6 that “claim | recites a specific and structured method for improving the functionality of a machine-learning-based cardiac diagnostic system through technical signal processing and classification workflows. The claim involves receiving multi-channel intracardiac electrogram (IEGM) signals, segmenting the data using temporal alignment across different electrode recordings, and extracting segment attributes for classification. The system applies a trained time series classifier to label each segment and generates a time series report that chronologically analyzes the segments and correlates observed patterns with clinical cardiac anomalies. These steps go beyond generic data analysis and reflect a concrete technical process that integrates synchronized signal acquisition, machine-learned pattern recognition, and medically relevant report generation. The claimed method addresses the technical problem of automating accurate, scalable cardiac event labeling in complex electrophysiological data streams.” Applicant is asserting the abstract idea itself as the improvement. However, the abstract idea cannot be an “additional element” that shows integration into a practical application. The order of calculations and the particular calculations claimed do not make the abstract idea any less abstract. The claims are currently structured as simply using a generic computer to implement the abstract idea (mental process), which is not enough to show a practical application.
Applicant then argues on page 7 that “Amended claim 1, like Claim 3 of USPTO Example 47, is patent eligible because it applies a judicial exception, namely, data analysis using machine learning, in a specific and structured manner that improves a technological process, namely, the automated labeling and interpretation of intracardiac electrogram (IEGM) data for clinical diagnostics. The steps in claim | provide a concrete and integrated solution to a technical problem: how to scale and standardize high-fidelity cardiac signal interpretation using synchronized machine learning- driven analysis. The claim reflects a practical application by tying the data-driven classification to real-world cardiac diagnostics, enabling consistent identification of events such as arrhythmias or procedural effects (e.g., ablation) across multiple electrodes and patients. By embedding temporal synchronization, classifier training, and report generation into a unified system, the claimed invention improves the accuracy, traceability, and utility of cardiac diagnostics in a way that cannot be practically performed by a human alone.” Examiner disagrees because each case turns on its own facts. Example 47 is directed to a different field of endeavor, in which the practical application is an improvement of the functioning of a computer which is seen in steps d-f. The instant claims fail to show the improvement of the functioning of a computer or in the technological field. The use of a processor and algorithms are simply automating the abstract idea on a computer.
Applicant then argues on page 8 that “claim 21 further the practical integration of the alleged abstract idea. Claim 21 recites “wherein receiving the plurality of time series data comprises prepossessing the plurality of time series data using a digital filtering technique comprising at least a Finite Impulse Response (FIR) filter configured to reduce noise while preserving temporal signal fidelity for accurate analysis of timing-dependent features in ECG and IEGM signals of the time series data”. Applicant asserts that this limitation reflects a specific, technical implementation of signal pre-processing that enhances the accuracy and reliability of downstream machine learning-based classification. FIR filtering is not a generic or abstract step but a well-established digital signal processing technique that is purposefully applied here to retain critical timing information from biomedical signals, information that is essential for correctly identifying events such as QRS complexes, P-waves, or atrial fibrillation triggers. This preprocessing step directly contributes to the claimed apparatus’s ability to generate high-integrity diagnostic outputs and exemplifies how the invention integrates any alleged abstract idea into a practical application through meaningful signal conditioning and model support.” Examiner disagrees because a filter is a mathematical concept using an equation for filtering high or low frequencies, allowing for the removal of noise.
Lastly, Applicant argues on page 9 that “the computing processor of claim | is not generic, either alone or in combination, because the claim recites a specific and structured set of operations that go beyond well-understood, routine, and conventional activities. Claim 1 describes an apparatus configured to receive intracardiac electrogram signals (IEGM) time series segments using temporal alignment across multiple electrode channels, extract segment attributes, and classify each segment using a trained time series classifier. The claim further requires generating a time series report that chronologically analyzes labeled segments and correlates event patterns to clinical cardiac anomalies. These coordinated steps reflect a non-conventional application of machine learning and signal processing, tailored to the technical challenge of automating accurate cardiac event interpretation across complex electrophysiological data streams. The integration of synchronized data acquisition, attribute extraction, classifier-based labeling, and diagnostic reporting enables a technical improvement to the functioning of a cardiac diagnostic system, not generic computing.” Examiner disagrees since the processing of data on a microcontroller unit is merely performing this process on a generic computer structure. The transmitting of signals is simply a generic computer function performed by a generic computer structure, wherein implementing the abstract idea with a generic computer is not enough to show integration into a practical application or significantly more than the abstract idea itself. The transmission of data to and from the sensor systems is merely data gathering, which is insignificant extra-solution activity. The ECG sensor, electrodes for IEGM are claimed very generically and are used only to gather the data they are designed for. These are well-understood, routine and conventional structure since the diagnostic art in Zhao et al (US 20170258356) teaches the use of ECG/EKG sensors to collect ECG signals ([0006]). The “multi-electrode catheter” is claimed very generically and are used only to gather the data they are designed for. These are well-understood, routine and conventional structure in the diagnostic art since Ravuna et al. (US 20210386355) teaches multi-electrode catheter electrodes [generic data gathering structure] for collecting electrograms ([0129]). Therefore, the rejection is maintained.
Applicant’s arguments, see pages 10-12, filed 07/28/2025, with respect to the rejection(s) of the claim(s) under 35 U.S.C. 103 have been fully considered and are persuasive. Amendments obviate the rejection of record. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of Ravuna et al. (US 20210386355) (IDS) (Hereinafter Ravuna) in view of Tzaikouri et al. (“Classification of AF and other arrhythmias from a short segment of ECG using dynamic time warping”, Computing in Cardiology 2017; VOL 44, 2017)(Hereinafter Tzaikouri), Liao et al. (“Deep Learning Classification of Unipolar Electrograms in Human Atrial Fibrillation: Application in Focal Source Mapping” Front Physiol. Jul 30, 2021)(Hereinafter Liao), and Moorman (US 20170127964) (IDS) (Hereinafter Moorman).
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.
Claim(s) 21 is/are 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.
Regarding claim 21, it is unclear how an FIR filter will be applied to ECG signals of the plurality of time series data of claims 1 and 11 because the independent claims are directed to IEGM and not ECG signals.
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.
Claims 1, 3-7, 9, 11, 13-17, and 19-21 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Each of independent claims 1 and 11 recites a step employs a temporal alignment technique to synchronize the segments across multiple channels of data…identify one or more segment attributes for each time series segment of the plurality of time series segments… classifying each time series segment of the plurality of time series segments to at least one time series label using the trained time series classifier, which is a mental process. This judicial exception is not integrated into a practical application because the generically recited computer elements (ie. a memory, a processor), determining values, and generating a labeled time series as a function of classification do not add a meaningful limitation to the abstract idea because they amount to simply implementing the abstract idea on a computer. The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional limitations are to receiving data, processing data, and generating a labeled time series as a function of classification, which are all well-understood, routine, and conventional computer functions. See MPEP § 2106.05(d).
MPEP 2106(III) outlines steps for determining whether a claim is directed to statutory subject
matter. The stepwise analysis for the instant claim is provided here.
Step 1 – Statutory categories
Claim 1 is directed to a system (i.e. machine) and thus meets the step 1 requirements.
Claim 11 is directed to a method and thus meets the step 1 requirements.
Step 2A – Prong 1 – Judicial exception (j.e.)
Regarding claims 1 and 11, the following step is an abstract idea:
“employs a temporal alignment technique to synchronize the segments across multiple channels of data…identify one or more segment attributes for each time series segment of the plurality of time series segments… classifying each time series segment of the plurality of time series segments to at least one time series label using the trained time series classifier”, which is a mental process when given its broadest reasonable interpretation. As discussed in MPEP 2106.04(a)(2)(II), the mental process grouping includes observations, evaluations, judgements, and opinions. In this case, a human could time-align different streams of data manually, identify segment attributes in each segment, and classify each segment.
Step 2A – Prong 2 – additional elements to integrate j.e. into a practical application
Regarding claims 1 and 11, the abstract idea is not integrated into a practical application.
The following claim elements do not add any meaningful limitation to the abstract idea:
- “a memory”, and “a processor” are recited at a high level of generality amounting to generic computer components for implementing abstract idea [MPEP 2106.05(b)];
It is noted that the classifier and the temporal alignment technique are by definition automating the human thinking process with a computer.
- “catheter” and “electrodes” of claim 3 is data gathering structures for the insignificant extra-solution activity of data gathering [MPEP 2106.05(b)];
- “time series data”, “plurality of time series segments”, “IEGM”, “time series label”, “labeled time series segment”, “report”, “chronological analysis”, “event patterns”, “correlation to clinical cardiac anomalies”, and “segment attributes” are data (gathering, selecting, and displaying) that is necessary to implement the abstract idea on a computer amounting to insignificant extra-solution activity [MPEP 2106.05(g)].
Step 2B – significantly more/inventive concept
Regarding claims 1 and 11, the abstract idea is not integrated into a practical application.
The following claim elements do not add any meaningful limitation to the abstract idea:
- “a memory”, and “a processor” are recited at a high level of generality amounting to generic computer components for implementing abstract idea [MPEP 2106.05(b)];
It is noted that the classifier and the temporal alignment technique are by definition automating the human thinking process with a computer.
- “catheter” and “electrodes” of claim 3 is data gathering structures for the insignificant extra-solution activity of data gathering [MPEP 2106.05(b)];
- “time series data”, “plurality of time series segments”, “IEGM”, “time series label”, “labeled time series segment”, “report”, “chronological analysis”, “event patterns”, “correlation to clinical cardiac anomalies”, and “segment attributes” are data (gathering, selecting, and displaying) that is necessary to implement the abstract idea on a computer amounting to insignificant extra-solution activity [MPEP 2106.05(g)].
The additional elements of claims 1 and 11, when considered separately and in combination, do not add significantly more (ie. an inventive concept) to the abstract idea. As discussed above with respect to the integration of the abstract idea into a practical application, processing circuitry, and storage devices, along with their associated functions, are recited at a high level of generality and simply amount to implementing the abstract idea on a computer. The ECG sensor, electrodes for IEGM are claimed very generically and are used only to gather the data they are designed for. These are well-understood, routine and conventional structure since the diagnostic art in Zhao et al (US 20170258356) teaches the use of ECG/EKG sensors to collect ECG signals ([0006]). The “multi-electrode catheter” is claimed very generically and are used only to gather the data they are designed for. These are well-understood, routine and conventional structure in the diagnostic art since Ravuna et al. (US 20210386355) teaches multi-electrode catheter electrodes [generic data gathering structure] for collecting electrograms ([0129]).
Dependent claims 3-7, 9, 13-17 and 19-21 do not integrate the abstract idea into a practical application and do not add significantly more to the abstract idea of claim 1 and 10. The dependent claim limitations are directed to the data processing (claims 2, 4-7, 9, 12, and 14-17, 19-20) and to generic gathering structure (claims 3 and 13), which are insignificant extra-solution activity and do not amount to more than what is well-understood, routine, and conventional. Claim 21 uses a filter, which is a mathematical concept implemented onto the processor.
In summary, claims 1, 3-7, 9, 11, 13-17, and 19-21 are directed to an abstract idea without significantly more and, therefore, are patent ineligible.
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.
Claim(s) 1, 3-5, 7, 9, 11, 13-15, 17, and 19-21 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ravuna et al. (US 20210386355) (IDS) (Hereinafter Ravuna) in view of Tzaikouri et al. (“Classification of AF and other arrhythmias from a short segment of ECG using dynamic time warping”, Computing in Cardiology 2017; VOL 44, 2017)(Hereinafter Tzaikouri), Liao et al. (“Deep Learning Classification of Unipolar Electrograms in Human Atrial Fibrillation: Application in Focal Source Mapping” Front Physiol. Jul 30, 2021)(Hereinafter Liao), and Moorman (US 20170127964) (IDS) (Hereinafter Moorman).
Regarding claims 1 and 11, Ravuna teaches An apparatus for labeling a plurality of time series data (Abstract “The system includes a processor comprising a machine learning algorithm configured to receive a first heartbeat” [0006] “Electrocardiograms (ECGs) and electrograms (EGMs) are examples of heart mapping.”), wherein the apparatus comprises:
at least a processor ([0009] “The system includes a processor comprising a machine learning algorithm configured to receive a first heartbeat at an identified cardiac spatial location including a first set of attributes information corresponding to the first heartbeat;”); and
a memory communicatively connected to the at least a processor ([0053] “processor 114 may be configured to store patient data, such as patient biometric data in memory 118”), wherein the memory contains instructions configuring the at least a processor to.
However, Ravuna does not teach a plurality of segments from a DTW used for identifying one or more attributes and classifying each segment based on the trained data. Tziakouri, in the same field of endeavor, teaches the detection of AF via feature extract using a classification model (Abstract), similar to the device of Ravuna, and further teaches receive a plurality of time series data (Pg. 1 right col. lines 13-15 “A test set consisting of 3,658 ECG recordings of similar lengths will be used for the final scoring but is not available to participants until after the completion of the Challenge.”);
generate a plurality of time series segments for each time series represented within the plurality of time series data (Pg. 1 right col. lines 21-24 “As a pre-processing step, we initially apply a Savinsky-Golay smoothing filter to the ECG segments. The algorithm then classifies the short ECG segments into the four categories using a three-stage classification scheme:”);
wherein the at least a processor employs a temporal alignment technique to synchronize the segments across multiple channels of data, including at least recordings from different electrodes, wherein the technique comprises stretching or compressing segments of the time series to output a temporal alignment (Pg. 2 right col. lines 4-9 “Dynamic Time Warping (DTW) is a technique for aligning two time-series of different length, using a nonlinear transformation. Given two time series P and Q of lengths i=1…N and i=1…M respectively, a N×M matrix is constructed where element (i,j) contains the distance of points d(pi, qj).” Examiner notes that the DTW stretches and compresses the segments for temporal alignment. Examiner further notes that the this can be done to different recording electrodes since each participants each had their own lead/electrode. );
identify one or more segment attributes for each time series segment of the plurality of time series segments (Pg. 2 left col. lines 28-30 “Two categories of features were used: two features are based on a template matching scheme employing Dynamic Time Warping.”);
classify each time series segment of the plurality of time series segments to at least one time series label as a function of the one or more segment attributes (Pg. 2 left col. lines 6-13 “segments labelled as either AF or Other rhythm are collapsed into one group, collectively labelled as Abnormal. Using seven features extracted from the ECG segments, a Support Vector Machine (SVM) classifier is trained which is used to classify segments as Normal or Abnormal. Segments which are determined to be Abnormal after this step proceed to the next classification step.”), wherein classifying each time series segment of the plurality of time series segments to at least one time series label comprises:
training a time series classifier using time series training data, wherein the time series training data comprises examples of segment attributes correlated to examples of time series labels (Pg. 2 left col. lines 6-13 “segments labelled as either AF or Other rhythm are collapsed into one group, collectively labelled as Abnormal. Using seven features extracted from the ECG segments, a Support Vector Machine (SVM) classifier is trained which is used to classify segments as Normal or Abnormal. Segments which are determined to be Abnormal after this step proceed to the next classification step.”); and
classifying each time series segment of the plurality of time series segments to at least one time series label using the trained time series classifier (Pg. 2 left col. lines 6-13 “segments labelled as either AF or Other rhythm are collapsed into one group, collectively labelled as Abnormal. Using seven features extracted from the ECG segments, a Support Vector Machine (SVM) classifier is trained which is used to classify segments as Normal or Abnormal. Segments which are determined to be Abnormal after this step proceed to the next classification step.”); and
generate at least one labeled time series segment for each time series segment of the plurality of time series segments as a function of the classification (Pg. 2 left col. lines 6-13 “segments labelled as either AF or Other rhythm are collapsed into one group, collectively labelled as Abnormal. Using seven features extracted from the ECG segments, a Support Vector Machine (SVM) classifier is trained which is used to classify segments as Normal or Abnormal. Segments which are determined to be Abnormal after this step proceed to the next classification step.” See AF classification at the end.) to increase accuracy and performance (Table 1). It would have been obvious to one skilled in the art, prior to the effective filing date of the claimed invention to modify the invention of Ravuna, with the plurality of segments from a DTW used for identifying one or more attributes and classifying each segment based on the trained data of Tziakouri, because such a modification would allow to increase accuracy and performance.
However, Tziakouri and Ravuna does not teach the plurality of time series data comprises a plurality of intracardiac electrogram (IEGM) signals and chronological analysis of intracardiac electrogram (IEGM) segments, identifying event patterns and their correlation to clinical cardiac anomalies. Liao, in the same field of endeavor, teaches diagnosing AF via the classification of segments (Abstract), and further teaches wherein the plurality of time series data comprises a plurality of intracardiac electrogram (IEGM) signals (Pg. 10 left col. lines 9-10 “To our knowledge, our study is the first application of DL to classify raw, intracardiac EGMs during AF.” Fig. 3 where EGM is split into periods for the patient. Page 5 left col. lines 16-18 “Among the 78 patients, a total of 13,184 periodic unpolar EGMs were recorded of which 1,220 (9.2%) had a dominant, sustained QS morphology (i.e., FaST) and the remaining 11,964 (90.7%) were non-FaST”) and chronological analysis of intracardiac electrogram (IEGM) segments, identifying event patterns and their correlation to clinical cardiac anomalies (Pg. 9 right col. lines 25-27 “These findings highlight the modest precision in the manual interpretation and classification of periodic unipolar QS EGMs during AF.”) to provide an alternative way of classifying AF invasively (Pg. 9 right col. lines 3-19). It would have been obvious to one skilled in the art, prior to the effective filing date of the claimed invention to modify the invention of Ravuna, with the plurality of time series data comprises a plurality of intracardiac electrogram (IEGM) signals and chronological analysis of intracardiac electrogram (IEGM) segments, identifying event patterns and their correlation to clinical cardiac anomalies of Liao, because such a modification would allow to provide an alternative way of classifying AF invasively.
However, Ravuna does not teach a times series report as a function of the one or more segment attributes. Moorman, in the same field of endeavor, teaches the classifying cardiac rhythms by segmenting into a plurality of segments and analysze the parameter data using algorithms (Abstract), similar to the device of Tziakouri and Ravuna, and further teaches generate a times series report as a function of the one or more segment attributes wherein the time series report comprises a … identifying event patterns and their correlation to clinical cardiac anomalies ([0072] “The average histograms for the 10-minute segments represented in FIG. 2 are illustrated in FIG. 7. As expected, in AF condition most templates find very few matches 704, whereas a bell-shape histogram distribution was obtained for the NSR series 702. In case of PVCs the histogram is similar to the AF case, due to the high amount of ventricular ectopic beats which produces high values in the first bins of the histogram 706. For SR with PACs 708, the histogram [report] reports a peak in correspondence of bin 5 and the majority of templates finds a number of matches less than 5.”) to report peaks of AF condition ([0072]). It would have been obvious to one skilled in the art, prior to the effective filing date of the claimed invention to modify the invention of Ravuna, with the times series report as a function of the one or more segment attributes of Moorman, because such a modification would allow to report peaks of AF condition.
Regarding claims 3 and 13, claims 1 and 11 are obvious over Tziakouri and Ravuna. Ravuna teaches wherein receiving the plurality of intracardiac electrograms signals comprises receiving the plurality of intracardiac electrograms signals from a catheter ([0129] “a multi-electrode catheter may be advanced into a chamber of the heart...Electrograms (EGMs) may be recorded from each of the electrodes in contact with a cardiac surface relative to a temporal reference such as the onset of the P-wave in sinus rhythm from a body surface ECG.”).
Regarding claims 4 and 14, claims 1 and 11 are obvious over Tziakouri and Ravuna. Tziakouri teaches wherein the plurality of time series data comprises a plurality of electrocardiogram (ECG) signals (Pg. 1 right col. lines 13-15 “A test set consisting of 3,658 ECG recordings of similar lengths will be used for the final scoring but is not available to participants until after the completion of the Challenge.”).
Regarding claims 5 and 15, claims 1 and 11 are obvious over Tziakouri and Ravuna. Tziakouri teaches wherein training the time series classifier comprises generating time series training data using a labeling module (Pg. 2 left col. lines 6-13 “segments labelled as either AF or Other rhythm are collapsed into one group, collectively labelled as Abnormal. Using seven features extracted from the ECG segments, a Support Vector Machine (SVM) classifier is trained which is used to classify segments as Normal or Abnormal. Segments which are determined to be Abnormal after this step proceed to the next classification step.”).
Regarding claims 7 and 17, claims 1 and 11 are obvious over Tziakouri and Ravuna. Tziakouri teaches wherein the memory further instructs the at least a processor to generate an attribute score as a function of the one or more segment attributes (Table 1 F1 score.).
Regarding claims 9 and 19, claims 1 and 11 are obvious over Tziakouri and Ravuna. Ravuna teaches wherein training the training a time series classifier comprises:
updating the time series training data as a function of an input and outputs of a previous time series classifier ([0267] “At 2324, a physician can manually fix the mapping annotation in the EP mapping system if the mapping annotation is not accurate. The machine learning system preferably receives data relating to the mapping annotation manually fixed by a physician in the EP mapping system.”); and
retraining the time series classifier using the updated time series training data ([0269] “2326, the machine learning system preferably further trains the deployed model by using new EP signals and the mapping annotation manually fixed by the physician in the EP mapping system.”).
Regarding claim 20, claims 1 and 11 are obvious over Tziakouri and Ravuna. Tziakouri further teach wherein the memory further instructs the at least a processor to perform a temporal alignment of the labeled time series segments (Pg. 2 right col. lines 4-9 “Dynamic Time Warping (DTW) is a technique for aligning two time-series of different length, using a nonlinear transformation. Given two time series P and Q of lengths i=1…N and i=1…M respectively, a N×M matrix is constructed where element (i,j) contains the distance of points d(pi, qj).” Examiner notes that the DTW stretches and compresses the segments for temporal alignment. Examiner further notes that the this can be done to different recording electrodes since each participants each had their own lead/electrode.).
Claim(s) 6 and 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ravuna et al. (US 20210386355) (IDS) (Hereinafter Ravuna) in view of Tzaikouri et al. (“Classification of AF and other arrhythmias from a short segment of ECG using dynamic time warping”, Computing in Cardiology 2017; VOL 44, 2017)(Hereinafter Tzaikouri) and Amrani et al. (“Very deep feature extraction and fusion for arrhythmias detection” S.I. : Deep Learning for Biomedical and Healthcare Applications Volume 30, pages 2047–2057, (2018)) (Hereinafter Amrani).
Regarding claims 6 and 16, claims 1 and 11 are obvious over Tziakouri and Ravuna. However, Ravuna does not teach standardizing to a canonical data format, verifying against pre-defined configuration, and annotating the verified examples of segment attributes. Amrani, in the same field of endeavor, teaches the use of ECG to diagnose arrhythmias, similar to Tziakouri and Ravuna, and further teaches wherein generating time series training data comprises:
standardizing each one of the examples of segment attributes into a canonical data format (Pg. 2050 left col. lines 22-25 “we propose multi-canonical correlation analysis (MCCA) to learn selective adaptive layer’s features such that the resulting representations are highly linearly correlated and therefore speed up the training task and improve the performance.”); and
verifying, for each one of the standardized examples of segment attributes, against a set of pre-defined configuration settings (Pg. 2050 right col. lines 5-11 “Suppose that X(p × n) and Y(q × n) are two matrices containing n training feature vectors. The aim of CCA is to find the projection direction of and that maximizes the pairwise correlations and across the two feature sets [35] as shown in Fig. 3.”); and
annotating, at the labeling module, the verified examples of segment attributes with the examples of time series labels to generate time series training data (Pg. 2051 left col. lines 7-13 “We assume that we have k sets of features , i = 1, 2,…,k which are sorted by their rank as: rank(F1) ≥ rank(F2) ≥ ≥ rank(Fk). MCCA applies CCA on two sets of features at the same time to attain the maximum possible length of the fused feature vector, of which in each step the two feature sets with the highest ranks are fused together as illustrated in Fig. 4.” The labeling of the rank for each are verified examples of the strongest correlation and therefore ranked highest.) to speed up the training task and improve the performance (Pg. 2050 left col. lines 22-25). It would have been obvious to one skilled in the art, prior to the effective filing date of the claimed invention to modify the invention of Ravuna, with the standardizing to a canonical data format, verifying against pre-defined configuration, and annotating the verified examples of segment attributes of Amrani, because such a modification would allow to speed up the training task and improve the performance.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Costa et al. (“Classification of Persistent Atrial Fibrillation Targets Using Machine Learning on Multipolar Electrograms”, https://hal.science/hal-05288084v1 Sept. 2025).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MOUSSA M HADDAD whose telephone number is (571)272-6341. The examiner can normally be reached M-TH 8:00-6:00.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Jennifer McDonald can be reached at (571) 270-3061. 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.
/MOUSSA HADDAD/Examiner, Art Unit 3796
/Jennifer Pitrak McDonald/Supervisory Patent Examiner, Art Unit 3796