Detailed Action
This action is in response to Applicant's communications filed 21 October 2024.
Claim(s) 1, 2, 4, 11, 12, and 14 was/were amended. Claim(s) 10 and 20 was/were cancelled. No claims were withdrawn. Claims 21-22 were added. Therefore, claims 1-9, 11-19, and 21-22 are pending in this Application.
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 .
Response to Amendments/Arguments
Applicant's arguments, filed 21 October 2024, regarding the rejections of claims 1-9, 11-19, and 21-22 under 35 USC 103have been fully considered but are not persuasive.
Applicant argues that the prior art references Rajpurkar, Kelwade, and Ghiasi (Remarks, p. 6) teach classifications of rhythms, as opposed to individual beats, as claimed. However, Rajpurkar discloses in its Related Works section that much work has been done in beat classification ("Much work has been done to automate the extraction of other features from the ECG. For example, beat classification is a common sub-problem of heart-arrhythmia classification…" sec. 6, p. 5). Additionally, Rajpurkar teaches classification on individual heartbeats based on the rhythms over time ("We identify 12 heart arrhythmias, sinus rhythm and noise for a total of 14 output classes. The arrhythmias are characterized by a variety of features. Table 2 in the Appendix shows an example of each rhythm type we classify. The noise label is assigned when the device is disconnected from the skin or when the baseline noise in the ECG makes identification of the underlying rhythm impossible. The morphology of the ECG during a single heartbeat as well as the pattern of the activity of the heart over time determine the underlying rhythm." sec. 3, p.3). The claims do not prohibit the use of the pattern of the activity of the heart over time when classifying heartbeats. Additionally, prior art of record Lin et al. (Heartbeat Classification Using Normalized RR Intervals and Morphological Features) teaches the classification of each individual heartbeat ("There are over 109,000 beats that are individually labeled as one of 15 possible heartbeat classes… The heartbeat classes included in this study are class N, consisting of the normal and bundle branch block beats, class S, consisting of supraventricular ectopic beats, and class V, consisting of ventricular ectopic beats, according to the AAMI recommendations." sec. 2.1, p. 2-3).
The rejection of the dependent claims for depending from rejected claims is maintained.
For the aforementioned reasons, claims 1-9, 11-19, and 21-22 are rejected under 35 USC 103.
Applicant’s arguments, filed 21 October 2024, with respect to the rejections of claims 1-9, 11-19, and 21-22 under 35 USC 103 are regarding newly amended or newly added claims and are addressed in the current rejection.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1, 5-9, 11, 15-19, and 21-22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Rajpurkar et al. (Cardiologist-Level Arrhythmia Detection with Convolutional Neural Networks, hereinafter "Rajpurkar"), Kelwade et al. (Prediction of Cardiac Arrhythmia using Artificial Neural Network, hereinafter "Kelwade"), Ghiasi et al. (Atrial Fibrillation Detection Using Feature Based Algorithm and Deep Convolutional Neural Network, hereinafter "Ghiasi"), and Lin et al. (Heartbeat Classification Using Normalized RR Intervals and Morphological Features, hereinafter "Lin").
Regarding Claim 1,
Rajpurkar teaches a computer-implemented method for classifying heartbeats using patient electrocardiogram (ECG) data, which comprises unstructured waveform data and structured time interval data, the method comprising: providing the unstructured waveform data ("The network takes as input a time-series of raw ECG signal" sec. 2, p. 2; "The morphology of the ECG during a single heart-beat as well as the pattern of the activity of the heart over time determine the underlying rhythm. In some cases the distinction can be subtle yet critical for treatment." sec. 3, p. 3) to a first neural network in a first path of a machine learning architecture (Figure 2; "We train a 34-layer convolutional neural network (CNN) to detect arrhythmias in arbitrary length ECG time-series." sec. 2, p. 2);
generating a first set of output values by using the first neural network to analyze the unstructured waveform data (Figure 2; "We use a convolutional neural network for the sequence-to-sequence learning task. The high-level architecture of the network is shown in Figure 2. The network takes as input a time-series of raw ECG signal, and outputs a sequence of label predictions. The 30 second long ECG signal is sampled at 200Hz, and the model outputs a new prediction once every second. We arrive at an architecture which is 33 layers of convolution followed by a fully connected layer and a softmax." sec. 2, p. 2).
Rajpurkar does not explicitly teach providing the structured time interval data to a second neural network in a second path of the machine learning architecture; and generating a second set of output values by using the second neural network to analyze the structured time interval data.
Kelwade teaches providing the structured time interval data to a second neural network in a second path of the machine learning architecture (Figure 8; "The Multi-Layer Perceptron (MLP) neural network is the most useful supervised neural network" sec. 3.6, p. 33) ("In this study of neural network for time series analysis, the heart rate time series i.e. RR interval time series the artificial neural network is used as a predictor to detect the cardiac arrhythmia into five heart diseases" sec. 4, p. 34); and
generating a second set of output values by using the second neural network to analyze the structured time interval data ("In this study of neural network for time series analysis, the heart rate time series i.e. RR interval time series the artificial neural network is used as a predictor to detect the cardiac arrhythmia into five heart diseases... The combination of linear parameters in frequency domains and non-linear parameters derived from the HRV signal are used as input features for the neural network." sec. 4, p. 34; these inputs for the neural network do not include the unstructured waveform data).
Rajpurkar and Kelwade are analogous art because both are directed to heartbeat classifiers. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classifier of Rajpurkar with the heartbeat classifier of Kelwade. The modification would have been obvious because one of ordinary skill in the art would be motivated to use classifiers with satisfactory performances in classifying different types of heartbeats, as suggested by Kelwade ("The proposed algorithm is predicted with next 30% of the datasets and satisfactory results obtained with prediction overall accuracy of 97%. " Abstract, p. 30; "The accuracy of discrimination of NSR, PVC, AFIB, LBBB and BII are 100%, 80%, 100%, 96.70%, and 100%, respectively." sec. 4, p. 34).
Rajpurkar does not explicitly teach generating a classification for each of the individual heartbeats based on the first set of output values and the second set of output values, wherein the first neural network comprises a different type of neural network than the second neural network.
Ghiasi teaches generating a classification […] ("classify ECG signals into 4 classes of arrhythmias and develop an automated algorithm to assess AF." p. 1) based on the first set of output values (Figure 1, Output 2) and the second set of output values (Figure 1, Output 1), wherein the first neural network (Figure 1, CNN) comprises a different type of neural network than the second neural network (Figure 1, Net) ("The outputs of both classifiers discussed in the previous section are combined using a decision table. (Table 3) Output1 and output2 refer to the prediction of neural network and CNN respectively." sec. 3, p. 3).
PNG
media_image1.png
297
643
media_image1.png
Greyscale
Rajpurkar and Ghiasi are analogous art because both are directed to heartbeat classification. It would have been obvious to one of ordinary skill in the art before the effective filing date to modify the heartbeat classifiers of the Rajpurkar/Kelwade to provide a combined classification. The modification would have been obvious because one of ordinary skill in the art would be motivated to distinguish the output from normal and noisy signals, as suggested by Ghiasi ("This research aims to distinguish different kinds of arrhythmias form normal and noisy signals." p. 4).
Rajpurkar does not explicitly teach generating a classification for each of the individual heartbeats based on the first set of output values and the second set of output values, wherein the classification is selected from one of the following: a normal beat or a ventricular beat.
Lin teaches generating a classification for each of the individual heartbeats based on the first set of output values and the second set of output values ("Figure 1 is a block diagram of the proposed heartbeat classification system. The purpose of the signal preprocessing is to remove the high-frequency noise signal and baseline drift using a second-order low-pass filter and two median filters, respectively. The heartbeat classification is based on the nonnormalized and normalized RR interval features, and the morphological features were extracted using the wavelet analysis and the linear prediction modeling and are performed by linear discriminant classification." sec. 2, p. 2), wherein the classification is selected from one of the following: a normal beat or a ventricular beat ("The extracted features described in the previous section for the classification of normal, supraventricular ectopic, and ventricular ectopic heartbeats are summarized in Table 2" sec. 3, p. 8).
Rajpurkar and Lin are analogous art because both are directed to heartbeat classification. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classification system of Rajpurkar/Kelwade/Ghiasi/Schwab combination with the combined heartbeat classification system of Lin. The modification would have been obvious because one of ordinary skill in the art would be motivated to improve global performance, as suggested by Lin ("the combination of the wavelet and linear prediction morphological features has higher global performance than only using the wavelet features or the linear prediction features." p. 1).
Regarding Claim 5,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 1. Rajpurkar further teaches wherein the first neural network comprises a convolutional neural network (Figure 2; "We train a 34-layer convolutional neural network (CNN) to detect arrhythmias in arbitrary length ECG time-series." sec. 2, p. 2).
Regarding Claim 6,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 5. Kelwade further teaches wherein the second neural network comprises a fully-connected neural network (Figure 8; "The Multi-Layer Perceptron (MLP) neural network is the most useful supervised neural network" sec. 3.6, p. 33; an MLP is a fully-connected neural network) ("In this study of neural network for time series analysis, the heart rate time series i.e. RR interval time series the artificial neural network is used as a predictor to detect the cardiac arrhythmia into five heart diseases" sec. 4, p. 34).
Rajpurkar and Kelwade are analogous art because both are directed to heartbeat classifiers. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classifier of Rajpurkar with the heartbeat classifier of Kelwade. The modification would have been obvious because one of ordinary skill in the art would be motivated to use classifiers with satisfactory performances in classifying different types of heartbeats, as suggested by Kelwade ("The proposed algorithm is predicted with next 30% of the datasets and satisfactory results obtained with prediction overall accuracy of 97%. " Abstract, p. 30; "The accuracy of discrimination of NSR, PVC, AFIB, LBBB and BII are 100%, 80%, 100%, 96.70%, and 100%, respectively." sec. 4, p. 34).
Regarding Claim 7,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 6. Kelwade further teaches wherein the second path does not comprise a convolution layer (Figure 8; "The Multi-Layer Perceptron (MLP) neural network is the most useful supervised neural network" sec. 3.6, p. 33) ("In this study of neural network for time series analysis, the heart rate time series i.e. RR interval time series the artificial neural network is used as a predictor to detect the cardiac arrhythmia into five heart diseases" sec. 4, p. 34).
Rajpurkar and Kelwade are analogous art because both are directed to heartbeat classifiers. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classifier of Rajpurkar with the heartbeat classifier of Kelwade. The modification would have been obvious because one of ordinary skill in the art would be motivated to use classifiers with satisfactory performances in classifying different types of heartbeats, as suggested by Kelwade ("The proposed algorithm is predicted with next 30% of the datasets and satisfactory results obtained with prediction overall accuracy of 97%. " Abstract, p. 30; "The accuracy of discrimination of NSR, PVC, AFIB, LBBB and BII are 100%, 80%, 100%, 96.70%, and 100%, respectively." sec. 4, p. 34).
Regarding Claim 8,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 1. Rajpurkar further teaches wherein the first neural network does not process the structured time interval data ("The network takes as input a time-series of raw ECG signal" sec. 2, p. 2; "The morphology of the ECG during a single heart-beat as well as the pattern of the activity of the heart over time determine the underlying rhythm. In some cases the distinction can be subtle yet critical for treatment." sec. 3, p. 3), Kelwade wherein the second neural network does not process the unstructured waveform data ("In this study of neural network for time series analysis, the heart rate time series i.e. RR interval time series the artificial neural network is used as a predictor to detect the cardiac arrhythmia into five heart diseases... The combination of linear parameters in frequency domains and non-linear parameters derived from the HRV signal are used as input features for the neural network." sec. 4, p. 34; these inputs for the neural network do not include the unstructured waveform data) .
Regarding Claim 9,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 1. Kelwade further teaches wherein the structured time interval data comprises R-R interval data ("In this study of neural network for time series analysis, the heart rate time series i.e. RR interval time series the artificial neural network is used as a predictor to detect the cardiac arrhythmia into five heart diseases... The combination of linear parameters in frequency domains and non-linear parameters derived from the HRV signal are used as input features for the neural network." sec. 4, p. 34; these inputs for the neural network do not include the unstructured waveform data).
Rajpurkar and Kelwade are analogous art because both are directed to heartbeat classifiers. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classifier of Rajpurkar with the heartbeat classifier of Kelwade. The modification would have been obvious because one of ordinary skill in the art would be motivated to use classifiers with satisfactory performances in classifying different types of heartbeats, as suggested by Kelwade ("The proposed algorithm is predicted with next 30% of the datasets and satisfactory results obtained with prediction overall accuracy of 97%. " Abstract, p. 30; "The accuracy of discrimination of NSR, PVC, AFIB, LBBB and BII are 100%, 80%, 100%, 96.70%, and 100%, respectively." sec. 4, p. 34).
Regarding Claim 21,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 1. Lin further teaches wherein the classification is further selected from a supra-ventricular beat ("The extracted features described in the previous section for the classification of normal, supraventricular ectopic, and ventricular ectopic heartbeats are summarized in Table 2" sec. 3, p. 8).
Rajpurkar and Lin are analogous art because both are directed to heartbeat classification. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classification system of Rajpurkar/Kelwade/Ghiasi/Schwab combination with the combined heartbeat classification system of Lin. The modification would have been obvious because one of ordinary skill in the art would be motivated to improve global performance, as suggested by Lin ("the combination of the wavelet and linear prediction morphological features has higher global performance than only using the wavelet features or the linear prediction features." p. 1).
Regarding Claim(s) 11, 15-19, and 22,
Claim(s) 11, 15-19, and 22 recite(s) a system including processor and memory storing instructions for performing functions corresponding to the method steps recited in claim(s) 1, 5-9, and 21, respectively. The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the limitations of claim(s) 11, 15-19, and 22 as set forth above in connection with claim(s) 1, 5-9, and 21. Therefore, claim(s) 11, 15-19, and 22 is/are rejected under the same rationale as respective claim(s) 1, 5-9, and 21.
Claim(s) 2-4, and 12-14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Rajpurkar et al. (Cardiologist-Level Arrhythmia Detection with Convolutional Neural Networks, hereinafter "Rajpurkar"), Kelwade et al. (Prediction of Cardiac Arrhythmia using Artificial Neural Network, hereinafter "Kelwade"), Ghiasi et al. (Atrial Fibrillation Detection Using Feature Based Algorithm and Deep Convolutional Neural Network, hereinafter "Ghiasi"), Lin et al. (Heartbeat Classification Using Normalized RR Intervals and Morphological Features, hereinafter "Lin"), and Schwab et al. (Beat by Beat: Classifying Cardiac Arrhythmias with Recurrent Neural Networks, hereinafter "Schwab").
Regarding Claim 2,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 1. Rajpurkar does not explicitly teach wherein the first set of output values and the second set of output values are inputted into a fully connected layer and concatenated using the fully connected layer in the machine learning architecture.
Schwab teaches wherein the first set of output values and the second set of output values are inputted into a fully connected layer and concatenated using the fully connected layer in the machine learning architecture (Figure 1, Level 2 Blender, p. 1; "Level 2 Blender. We use blending [22] to combine the predictions of our level 1 models and a set of ECG-wide features into a final per-class classification score. The additional features are the RWE and WE over the whole ECG and the absolute average deviation (AAD) of the WE and δRR of all beats. We employ a MLP with a softmax output layer as our level 2 blender model. In order to avoid overfitting to the training set, we train the MLP on the validation set." p. 2).
Rajpurkar and Schwab are analogous art because both are directed to heartbeat classification. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classifier of the Rajpurkar/Kelwade/Ghiasi/Lin combination with the classification metho of Schwab. The modification would have been obvious because one of ordinary skill in the art would be motivated to distinguish between distinguish between different types of signals, as suggested by Schwab ("We utilise an annotated dataset of 12,186 single-lead ECG recordings to build a diverse ensemble of recurrent neural networks (RNNs) that is able to distinguish between normal sinus rhythms, atrial fibrillation, other types of arrhythmia and signals that are too noisy to interpret. ").
Regarding Claim 3,
The Rajpurkar/Kelwade/Ghiasi/Lin/Schwab combination teaches the computer-implemented method of claim 2.
Lin further teaches wherein the fully connected layer outputs a probability ("posterior probability P" sec. 2.6, p. 7) for each heartbeat classification ("The extracted features described in the previous section for the classification of normal, supraventricular ectopic, and ventricular ectopic heartbeats are summarized in Table 2" sec. 3, p. 8).
Rajpurkar and Lin are analogous art because both are directed to heartbeat classification. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classification system of Rajpurkar/Kelwade/Ghiasi/Lin/Schwab combination with the combined heartbeat classification system of Lin. The modification would have been obvious because one of ordinary skill in the art would be motivated to improve global performance, as suggested by Lin ("the combination of the wavelet and linear prediction morphological features has higher global performance than only using the wavelet features or the linear prediction features." p. 1).
Regarding Claim 4,
The Rajpurkar/Kelwade/Ghiasi/Lin combination teaches the computer-implemented method of claim 2. Rajpurkar does not explicitly teach wherein the generating the classification comprises: verifying, using the machine learning architecture, that the heartbeat comprises a beat, based on concatenation of the first set of output values and the second set of output values.
Schwab teaches wherein the generating the classification comprises: verifying, using the machine learning architecture, that the heartbeat comprises a beat, based on concatenation of the first set of output values and the second set of output values (Figure 1, Level 2 Blender, Classification, Normal, AF, Other, Noise, p. 1; the classification of the heartbeats include beats such as "normal", "AF", and "Other", and nonbeats, such as "Noise").
Rajpurkar and Schwab are analogous art because both are directed to heartbeat classification. It would have been obvious to one of ordinary skill in the art before the effective filing date to combine the heartbeat classifier of the Rajpurkar/Kelwade/Ghiasi/Lin combination with the classification metho of Schwab. The modification would have been obvious because one of ordinary skill in the art would be motivated to distinguish between distinguish between different types of signals, as suggested by Schwab ("We utilise an annotated dataset of 12,186 single-lead ECG recordings to build a diverse ensemble of recurrent neural networks (RNNs) that is able to distinguish between normal sinus rhythms, atrial fibrillation, other types of arrhythmia and signals that are too noisy to interpret. ").
Regarding Claim(s) 12-14,
Claim(s) 12-14 recite(s) a system including processor and memory storing instructions for performing functions corresponding to the method steps recited in claim(s) 2-4, respectively. The Rajpurkar/Kelwade/Ghiasi/Lin/Schwab combination teaches the limitations of claim(s) 12-14 as set forth above in connection with claim(s) 2-4. Therefore, claim(s) 12-14 is/are rejected under the same rationale as respective claim(s) 2-4.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHARLES C KUO whose telephone number is (571)270-7477. The examiner can normally be reached M-F: 9:00 a.m. - 6:00 p.m..
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, David Yi can be reached on (571) 270-7519. 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.
/CHARLES C KUO/Examiner, Art Unit 2126
/VAN C MANG/Primary Examiner, Art Unit 2126
Prosecution Insights