# US 2019/0313920 A1 · System and method for classification of coronary artery disease based on metadata and cardiovascular signals

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0313920 A1
- Document ID: US20190313920A1
- Status: Published application
- Kind: A1
- Application number: 16/285,519
- Filed: 2019-02-26
- Published: 2019-10-17
- Earliest claimed: 2018-03-16
- Assignees: Tata Consultancy Services Limited
- Applicants: Tata Consultancy Services Limited
- Inventors: Rohan BANERJEE; Sakyajit BHATTACHARYA; Soma BANDYOPADHYAY; Arpan PAL; Kayapanda Muthana MANDANA
- Web page: https://patentgrep.com/patent/US20190313920A1

## CPC

- A61B 5/02007: {Evaluating blood vessel condition, e.g. elasticity, compliance}
- A61B 5/025: within occluders, e.g. responsive to Korotkoff sounds
- A61B 5/7264: {Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems}
- A61B 5/7267: {involving training the classification device}
- A61B 5/7275: {Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor}
- G16H 50/30: for calculating health indices; for individual health risk assessment
- A61B 5/02416: {using photoplethysmograph signals, e.g. generated by infrared radiation (A61B5/14552 takes precedence)}
- A61B 5/7246: {using correlation, e.g. template matching or determination of similarity}
- A61B 5/725: {using specific filters therefor, e.g. Kalman or adaptive filters (specific diagnostics methods using using bioelectric or biomagnetic signals A61B5/316)}
- A61B 5/7257: {using Fourier transforms}
- A61B 7/04: Electric stethoscopes
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- G16H 50/70: for mining of medical data, e.g. analysing previous cases of other patients
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

Non-invasive methods for accurately classifying Coronary Artery Disease (CAD) is a challenging task. In the present disclosure, a two stage classification is performed. In the first stage of classification, a metadata based rule engine is utilized to classify a subject into one of a confirmed CAD subject, a CAD subject and a non-CAD subject. Here, a set of optimal parameters are selected from a set of metadata associated with the subject based on a difference in frequency of occurrence of the CAD among a disease population and a non-disease population. Further, an optimal threshold associated with each optimal parameter is calculated based on an inflexion based correlation analysis. Further, the CAD subject, classified by the metadata based rule engine is further reclassified in a second stage by utilizing a set of cardiovascular signal into one of the CAD subject and the non-CAD subject.

## Claims

### Claim 1

A method for classification of Coronary Artery Disease (CAD) based on metadata and cardiovascular signals, the method comprising:
receiving, by a one or more hardware processors, a set of metadata associated with a subject and a set of cardiovascular signals associated with the subject, wherein each metadata from the set of meta data is associated with a value;
selecting, by the one or more hardware processors, a set of optimal parameters from the set of metadata based on a difference in frequency of occurrence of CAD among a disease population and a non-disease population;
computing, by the one or more hardware processors, an optimal threshold for each optimal parameter from the set of optimal parameters based on an inflexion point based correlation analysis;
classifying, by the one or more hardware processors, the subject by utilizing a metadata based rule engine into a category among a plurality of categories, wherein the plurality of categories comprising a confirmed CAD subject, a CAD subject and a non-CAD subject, wherein the metadata based rule engine is constructed by utilizing the set of optimal parameters and the optimal threshold associated with each optimal parameter; and
reclassifying, by the one or more hardware processors, the CAD subject into one of the confirmed CAD subject and the non-CAD subject based on a combination of a PPG signal classifier and a PCG signal classifier in parallel.

### Claim 2 (depends on 1)

The method as claimed in claim 1, wherein identifying, the optimal threshold for each optimal parameter from the set of optimal parameters based on the inflexion point based correlation analysis further comprising:
calculating a set of correlation coefficient for each value associated with each optimal parameter from the set of optimal parameters based on Bhattacharya distance between a histogram associated with the disease population and a histogram associated with the non-disease population;
plotting a two dimensional chart for each optimal parameter from the set of optimal parameters, wherein, the set of values associated with each optimal parameter is plotted along an X plane and the set of correlation coefficients associated with each optimal parameter is plotted along a Y plane; and
identifying the optimal threshold for each optimal parameter by selecting a highest inflexion point from a set of points associated with the two dimensional chart corresponding to each optimal parameter from the set of optimal parameters.

### Claim 3 (depends on 1)

The method as claimed in claim 1, wherein reclassifying, the CAD subject into one of the confirmed CAD subject and the non-CAD subject based on the combination of the PPG signal classifier and the PCG signal classifier in parallel further comprising:
receiving the set of cardiovascular signal associated with the CAD subject, wherein the set of cardiovascular signal comprises a PPG signal and a PCG signal;
extracting a set of PPG signal parameters from the PPG signal and a set of PCG signal parameters from the PCG signal in parallel;
calculate a first confidence score and a second confidence score in parallel, wherein the first confidence score is obtained by utilizing the PPG signal classifier and the second confidence score is obtained by utilizing the PCG signal classifier; and
reclassifying the CAD subject into one of the confirmed CAD subject and the non-CAD subject by selecting a highest confidence score among the first confidence score and the second confidence score.

### Claim 4 (depends on 1)

The method as claimed in claim 1, wherein the set of metadata comprising a set of demographic data and a set of clinical information.

### Claim 5 (depends on 3)

The method as claimed in claim 3, wherein the set of PCG signal parameters comprises a spectral power ratio, a spectral centroid, a spectral roll-off, a spectral flux, a kurtosis of time windows, a natural entropy, and a Tsallis entropy.

### Claim 6 (depends on 3)

The method as claimed in claim 3, wherein the set of PPG signal parameters comprises a cycle duration, a systolic upstroke time, a diastolic time, a trough to notch time, a notch to trough time, a peak to notch time and a pulse width.

### Claim 7 (depends on 3)

The method as claimed in claim 3, wherein the PPG signal classifier is trained by utilizing a PPG signal training data, wherein, the PPG signal training data includes the set of PPG signal parameters associated with the disease population and the set of PPG signal parameters associated with the non-disease population.

### Claim 8 (depends on 3)

The method as claimed in claim 3, wherein the PCG signal classifier is trained by utilizing a PCG signal training data, wherein, the PCG signal training data includes the set of PCG signal parameters associated with the disease population and the set of PCG signal parameters associated with the non-disease population.

### Claim 9

A system for classification of Coronary Artery Disease (CAD) based on metadata and cardiovascular signals, the system comprising:
one or more memories comprising programmed instructions and a repository for storing a set of metadata associated with a subject and a set of cardiovascular signal associated with the subject;
one or more hardware processors operatively coupled to the one or more memories, wherein the one or more hardware processors are capable of executing the programmed instructions stored in the one or more memories;
a PPG signal capturing device, a PCG signal capturing device and
a CAD analysis unit, wherein the CAD analysis unit is configured to:
receive, a set of metadata associated with a subject and a set of cardiovascular signal associated with the subject, wherein each metadata from the set of meta data is associated with a value;
select, a set of optimal parameters from the set of metadata based on a difference in frequency of occurrence of CAD among a disease population and a non-disease population;
compute, an optimal threshold for each optimal parameter from the set of optimal parameters based on an inflexion point based correlation analysis;
classify, the subject by utilizing a metadata based rule engine into a category among a plurality of categories, wherein the plurality of categories comprising a confirmed CAD subject, a CAD subject and a non-CAD subject, wherein the metadata based rule engine is constructed by utilizing the set of optimal parameters and the optimal threshold associated with each optimal parameter; and
reclassify, the CAD subject into one of the confirmed CAD subject and the non-CAD subject based on a combination of a PPG signal classifier and a PCG signal classifier in parallel.

### Claim 10 (depends on 9)

The system as claimed in claim 9, wherein the CAD analysis unit is configured to identify the optimal threshold for each optimal parameter from the set of optimal parameters based on the inflexion point based correlation analysis by:
calculating a set of correlation coefficient for each value associated with each optimal parameter from the set of optimal parameters based on Bhattacharya distance between a histogram associated with the disease population and a histogram associated with the non-disease population;
plotting a two dimensional chart for each optimal parameter from the set of optimal parameters, wherein, the set of values associated with each optimal parameter is plotted along an X plane and the set of correlation coefficients associated with each optimal parameter is plotted along a Y plane; and
identifying the optimal threshold for each optimal parameter by selecting a highest inflexion point from a set of points associated with the two dimensional chart corresponding to each optimal parameter from the set of optimal parameters.

### Claim 11 (depends on 9)

The system as claimed in claim 9, wherein the CAD analysis unit is configured to reclassify the CAD subject into one of the confirmed CAD subject and the non-CAD subject based on the combination of the PPG signal classifier and the PCG signal classifier in parallel by:
receiving the set of cardiovascular signal associated with the CAD subject, wherein the set of cardiovascular signal comprises a PPG signal and a PCG signal;
extracting a set of PPG signal parameters from the PPG signal and a set of PCG signal parameters from the PCG signal in parallel;
calculate a first confidence score and a second confidence score in parallel, wherein the first confidence score is obtained by utilizing the PPG signal classifier and the second confidence score is obtained by utilizing the PCG signal classifier; and
reclassifying the CAD subject into one of the confirmed CAD subject and the non-CAD subject by selecting a highest confidence score among the first confidence score and the second confidence score.

### Claim 12 (depends on 9)

The system as claimed in claim 9, wherein the set of metadata comprising a set of demographic data and a set of clinical information.

### Claim 13 (depends on 11)

The system as claimed in claim 11, wherein the set of PCG signal parameters comprises a spectral power ratio, a spectral centroid, a spectral roll-off, a spectral flux, a kurtosis of time windows, a natural entropy, and a Tsallis entropy.

### Claim 14 (depends on 11)

The system as claimed in claim 11, wherein the set of PPG signal parameters comprises a cycle duration, a systolic upstroke time, a diastolic time, a trough to notch time, a notch to trough time, a peak to notch time and a pulse width.

### Claim 15 (depends on 11)

The system as claimed in claim 11, wherein the PPG signal classifier is trained by utilizing a PPG signal training data, wherein, the PPG signal training data includes the set of PPG signal parameters associated with the disease population and the set of PPG signal parameters associated with the non-disease population.

### Claim 16 (depends on 11)

The system as claimed in claim 11, wherein the PCG signal classifier is trained by utilizing a PCG signal training data, wherein, the PCG signal training data includes the set of PCG signal parameters associated with the disease population and the set of PCG signal parameters associated with the non-disease population.

### Claim 17

One or more non-transitory machine readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors causes the one or more hardware processor to perform a method for classification of Coronary Artery Disease (CAD) based on metadata and cardiovascular signals, said method comprising:
receiving, by the one or more hardware processors, a set of metadata associated with a subject and a set of cardiovascular signals associated with the subject, wherein each metadata from the set of meta data is associated with a value;
selecting, by the one or more hardware processors, a set of optimal parameters from the set of metadata based on a difference in frequency of occurrence of CAD among a disease population and a non-disease population;
computing, by the one or more hardware processors, an optimal threshold for each optimal parameter from the set of optimal parameters based on an inflexion point based correlation analysis;
classifying, by the one or more hardware processors, the subject by utilizing a metadata based rule engine into a category among a plurality of categories, wherein the plurality of categories comprising a confirmed CAD subject, a CAD subject and a non-CAD subject, wherein the metadata based rule engine is constructed by utilizing the set of optimal parameters and the optimal threshold associated with each optimal parameter; and
reclassifying, by the one or more hardware processors, the CAD subject into one of the confirmed CAD subject and the non-CAD subject based on a combination of a PPG signal classifier and a PCG signal classifier in parallel.

### Claim 18 (depends on 17)

The one or more non-transitory machine readable information storage mediums of claim 17, wherein identifying, the optimal threshold for each optimal parameter from the set of optimal parameters based on the inflexion point based correlation analysis further comprising:
calculating a set of correlation coefficient for each value associated with each optimal parameter from the set of optimal parameters based on Bhattacharya distance between a histogram associated with the disease population and a histogram associated with the non-disease population;
plotting a two dimensional chart for each optimal parameter from the set of optimal parameters, wherein, the set of values associated with each optimal parameter is plotted along an X plane and the set of correlation coefficients associated with each optimal parameter is plotted along a Y plane; and
identifying the optimal threshold for each optimal parameter by selecting a highest inflexion point from a set of points associated with the two dimensional chart corresponding to each optimal parameter from the set of optimal parameters.

### Claim 19 (depends on 17)

The one or more non-transitory machine readable information storage mediums of claim 17, wherein reclassifying, the CAD subject into one of the confirmed CAD subject and the non-CAD subject based on the combination of the PPG signal classifier and the PCG signal classifier in parallel further comprising:
receiving the set of cardiovascular signal associated with the CAD subject, wherein the set of cardiovascular signal comprises a PPG signal and a PCG signal;
extracting a set of PPG signal parameters from the PPG signal and a set of PCG signal parameters from the PCG signal in parallel;
calculate a first confidence score and a second confidence score in parallel, wherein the first confidence score is obtained by utilizing the PPG signal classifier and the second confidence score is obtained by utilizing the PCG signal classifier; and
reclassifying the CAD subject into one of the confirmed CAD subject and the non-CAD subject by selecting a highest confidence score among the first confidence score and the second confidence score.

### Claim 20 (depends on 17)

The one or more non-transitory machine readable information storage mediums of claim 17, wherein the set of metadata comprising a set of demographic data and a set of clinical information.
