FORM 2
THE PATENTS ACT, 1970 (39 of 1970) & THE PATENTS RULES, 2003
COMPLETE SPECIFICATION (See section 10, rule 13) 1. Title of the invention: METHOD AND SYSTEM FOR IDENTIFYING CORONARY
ARTERY DISEASE PATIENTS USING PHOTOPLETHYSMOGRAM SIGNAL
2. App lica nt(s)
NAME NATIONALITY ADDRESS
TATA CONSULTANCY Indian Nirmal Building, 9th Floor,
SERVICES LIMITED Nariman Point, Mumbai- 400021,
Maharashtra, India
3. Preamble to the description
COMPLETE SPECIFICATION
The following specification particularly describes the invention and the manner in which it
is to be performed.

FIELD OF THE INVENTION
The present application generally relates to biomedical signal processing. More particularly, the application provides a method and system for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
BACKGROUND
[001] Huge number of IoT devices are available to promote health care management and wellness. It is undoubted that IoT Healthcare solutions can provide remote monitoring to support patients suffering from various diseases and disorders. But, a gamut of expensive sensor devices, sophisticated, periodic setup, maintenance and calibration as well as up-to-date training are required for such purpose to come to fruition. In order to promote widespread usage and affordability, such costly and extensive intricacies do not work positively towards the ubiquity and success of mobile and preventable health care, specifically in developing countries. As a result, deriving various physiological parameters of a person in a noninvasive and affordable manner is a significant task and challenge.
[002] But there are prior art which suggests the possibility of the analysis of various pathological conditions related to the cardio-vascular system from non-invasive physiological signal analysis techniques. In particular, there are several instances of the use of signals such as PPG and ECG. However, there are quite a few limitations in the prior art. None of the prior is directly and exactly related to coronary artery disease (CAD) detection from a physiological signal. They either broadly talk about the possible diagnosis of cardio-vascular diseases from such signals or are focused on the diagnosis of peripheral arterial disease (PAD). None of the above prior art have used heart sounds as an input to the diagnosis system. None of the prior art have talked about the fusion of different decisions for CAD diagnosis. Thereby, identifying coronary artery disease (CAD) patients by fusing the decisions of multiple classifier systems based on multiple physiological signals is still considered to be one of the biggest challenges of the technical domain.

OBJECTIVES OF THE INVENTION
[003] In accordance with the present invention, an objective is to provide for a method and system for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
[004] Another objective of the invention is to identify coronary artery disease (CAD) patients by fusing the decisions of multiple classifier systems based on multiple physiological signals.
[005] Yet another objective of the invention is to use low resolution photoplethysmogram (PPG) signal to find markers for coronary artery disease (CAD) patients.
SUMMARY OF THE INVENTION
[006] Before the present methods, systems, and hardware enablement are described, it is to be understood that this invention is not limited to the particular systems, and methodologies described, as there can be multiple possible embodiments of the present invention which are not expressly illustrated in the present disclosure. It is also to be understood that the terminology used in the description is for the purpose of describing the particular versions or embodiments only, and is not intended to limit the scope of the present invention which will be limited only by the appended claims.
[007] The present disclosure envisages a method and system which can identify coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
[008] In an embodiment of the invention, a method for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal is provided. The method comprises extracting photoplethysmogram (PPG) signals from a patient, determining PPG peak rate per minute (PPGPPM) from the extracted photoplethysmogram (PPG) signals and extracting of time and frequency domain features from the extracted photoplethysmogram (PPG) signals, selecting most matching mother wavelet using PPGPPM with respect to a particular photoplethysmogram

(PPG) signal window, decomposing the PPGPPM signal to extract wavelet features from the extracted photoplethysmogram (PPG) signals, computing the extracted features and selection of the most relevant statistical and entropy related features and classifying coronary artery disease (CAD) patients and non-coronary artery disease (CAD) patients by using the selected most relevant statistical and entropy features.
[009] In another embodiment of the invention, a system for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal is provided. The system (200) comprises of an image capturing device (202) coupled with a mobile communication device (204), a feature extraction module (206), a wavelet selection module (208), a wavelet decomposition module (210), a feature computation module (212) and a classification module (214).
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The foregoing summary, as well as the following detailed description of preferred embodiments, are better understood when read in conjunction with the appended drawings. For the purpose of illustrating the invention, there is shown in the drawings exemplary constructions of the invention; however, the invention is not limited to the specific methods and system disclosed. In the drawings:
[0011] Figure 1 shows a flow chart illustrating method for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
[0012] Figure 2 shows a block diagram of a system for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
[0013] Figure 3 illustrates photoplethysmogram (PPG) peak detection for a sample coronary artery disease (CAD) patient.

[0014] Figure 4 illustrates photoplethysmogram (PPG) peak detection for a sample non-coronary artery disease (non-CAD) patient.
[0015] Figure 5 illustrates the time domain features as extracted from the photoplethysmogram (PPG) signal.
[0016] Figure 6 illustrates a sample PPGPPM signal for a CAD patient.
[0017] Figure 7 illustrates a sample PPGPPM signal for a non-CAD patient.
[0018] Figure 8 illustrates the db3 mother wavelet in accordance with an embodiment of the invention.
DETAILED DESCRIPTION OF THE INVENTION
[0019] Some embodiments of this invention, illustrating all its features, will now be discussed in detail.
[0020] The words "comprising," "having," "containing," and "including," and other forms thereof, are intended to be equivalent in meaning and be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items, or meant to be limited to only the lis ted item or items.
[0021] It must also be noted that as used herein and in the appended claims, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Although any systems and methods similar or equivalent to those described herein can be used in the practice or testing of embodiments of the present invention, the preferred, systems and methods are now described. In the following description for the purpose of explanation and understanding reference has been made to numerous embodiments for which the intent is not to limit the scope of the invention.

[0022] One or more components of the invention are described as module for the understanding of the specification. For example, a module may include self-contained component in a hardware circuit comprising of logical gate, semiconductor device, integrated circuits or any other discrete component. The module may also be a part of any software programme executed by any hardware entity for example processor. The implementation of module as a software programme may include a set of logical instructions to be executed by a processor or any other hardware entity.
[0023] The disclosed embodiments are merely exemplary of the invention, which may be embodied in various forms.
[0024] The elements illustrated in the Figures interoperate as explained in more detail below. Before setting forth the detailed explanation, however, it is noted that all of the discussion below, regardless of the particular implementation being described, is exemplary in nature, rather than limiting. For example, although selected aspects, features, or components of the implementations are depicted as being stored in memories, all or part of the systems and methods consistent with the natural disaster prediction system and method may be stored on, distributed across, or read from other machine-readable media.
[0025] Method steps of the invention may be performed by one or more computer processors executing a program tangibly embodied on a computer-readable medium to perform functions of the invention by operating on input and generating output. Suitable processors include, by way of example, both general and special purpose microprocessors. Generally, the processor receives (reads) instructions and data from a memory (such as a read-only memory and/or a random access memory) and writes (stores) instructions and data to the memory. Storage devices suitable for tangibly embodying computer program instructions and data include, for example, all forms of non-volatile memory, such as semiconductor memory devices, including EPROM, EEPROM, and flash memory devices; magnetic disks such as internal hard disks and removable disks; magneto-optical disks; and CD-ROMs. Any of the foregoing may be supplemented by, or incorporated in, specially-designed ASICs (application-specific integrated circuits) or FPGAs (Field-Programmable Gate Arrays). A computer can generally also receive (read) programs and

data from, and write (store) programs and data to, a non-transitory computer-readable storage medium such as an internal disk (not shown) or a removable disk.
[0026] The present disclosure provides a method and system for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
[0027] Referring to Figure 1, it is a flow chart illustrating a method for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
[0028] The process starts at step 102, photoplethysmogram (PPG) signals are extracted from a patient using the image capturing device (202) coupled with the mobile communication device (204). At the step 104, PPG peak rate per minute (PPGPPM) is determined from the extracted photoplethysmogram (PPG) signals and time and frequency domain features are extracted from the extracted photoplethysmogram (PPG) signals. At step 106, the most matching mother wavelet is selected using PPG peak rate per minute (PPGPPM) with respect to a particular photoplethysmogram (PPG) signal window. At step 108, the PPGPPM signal is decomposed to extract wavelet features from the extracted photoplethysmogram (PPG) signals. At step 110, the extracted features are computed and the most relevant statistical and entropy related features are selected. The process end at step 112, coronary artery disease (CAD) patients and non-coronary artery disease (CAD) patients are classified by using the selected most relevant statistical and entropy features.
[0029] Referring to Figure 2, it is a block diagram illustrating system architecture for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal.
[0030] In another embodiment of the present invention, a system (200) for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal comprises of an image capturing device (202) coupled with a mobile communication device (204), a feature extraction module (206), a wavelet selection module (208), a wavelet decomposition module (210), a feature computation module (212) and a decision module (214).

[0031] In another embodiment of the present invention, the image capturing device (202) coupled with the mobile communication device (204) is adapted for extracting photoplethysmogram signals from the patient. The photoplethysmogram signals are extracted from patient’s peripheral body parts selected from a group comprising but not limited to finger, ear, and toe. In a specific embodiment, the photoplethysmogram signals are extracted from user’s forehead. The mobile communication device (204) captures photoplethysmogram signal in reflective mode. The mobile communication device (204) is selected from a group comprising of smart phone, mobile phone, laptop, tablet, and personal digital assistant.
[0032] The image capturing device (202) coupled with the mobile communication device (204) is a camera and have a light emitting source for extracting photoplethysmogram signals from the patient’s peripheral body parts selected from a group comprising but not limited to finger, ear, toe; forehead, thereby, obtaining a video sequence of the light, reflected from patient’s peripheral body parts.
[0033] PPG signal, as obtained using noninvasive ways through wearable devices and smartphones, is then used to compute the PPG peak rate-sampled once per second. The PPG peak rate also has a small storage footprint. The present invention proposes a method of using PPG signal to determine peaks and decompose the peak rate signal in time scale domain of the extracted PPG signals by applying wavelet transform and compute several statistical and entropy related features to find patterns to differentiate CAD and non CAD patients.
[0034] In another embodiment of the present invention, the extracted PPG signal is fed to a band-pass filter to remove any frequency component beyond 0.5 -10 Hz. Subsequently, the signal is shifted to same baseline to remove the potential error that could appear in feature extraction.
[0035] In another embodiment of the present invention, the feature extraction in the extracted PPG signal is performed on three levels- time domain feature extraction, frequency domain feature extraction and wavelet feature extraction. The classification between CAD and non CAD patients is performed based on said time domain feature extraction, frequency domain feature

extraction and wavelet feature extraction and the results of said three classifications are fused to decide on the confirmed classification of CAD and non CAD patients.
[0036] In another embodiment of the present invention, in the feature extraction module (206), PPG peak rate per minute (PPGPPM) signals is determined from the extracted photoplethysmogram (PPG) signals and time and frequency domain features are extracted from the extracted photoplethysmogram (PPG) signals. The number of peaks per second is calculated by reciprocating the each peak to next peak interval. Number of peaks per second value is obtained for each peak interval, where the peak intervals are spaced non-uniformly. Number of peaks per second value is resampled at uniform spacing, taking one sample per second and number of PPG peaks per minute (PPGPPM) is interpolated from number of peaks per second.
[0037] Figure 3 shows PPG peak detection for a sample CAD patient and Figure 4 shows PPG
peak detection for a sample non-CAD patient. In another embodiment of the invention,
once the peaks are detected, time domain PPG features are extracted from every cycle of the
PPG signal. The feature set, although not necessarily only containing the ones as specified
herein, includes the standard deviation of Tc over a window, Ts/Tc, Td/Tc and Ts/Td where
Tc : cycle duration
Ts: systolic upstroke time
Td: diastolic time
The time domain features are shown in Figure 5.
[0038] In another embodiment of the invention, time domain features are used to differentiate CAD and non-CAD patients using Support Vector Machine (SVM) based machine learning technique. The number of peaks value, obtained from the extracted PPG signal, is very close to the pulse rate and it helps to determine the Cardiac Cycle, Heart Rate, and Heart Rate Variability of a person. Sample PPGPPM signal is shown for CAD patient and in Figure 6 and sample PPGPPM signal is shown for non- CAD patient in Figure 7.
[0039] In another embodiment of the invention, in the wavelet selection module (208), the peaks per minute signal (PPGPPM), when plotted against time resolution of once a second, results in a signal plot that can be visually matched to a particular mother wavelet for a well-defined time

window. Based on this visual similarity, the mother wavelet is chosen for that time window of the signal. In an exemplary embodiment, after analyzing PPG peaks per minute from diverse data sources pertaining to several age group, genders and demographics, it has been observed that “db3” mother wavelet is most fit for this signal. The db3 mother wavelet is shown in Figure 8. In a particular setup, if the PPGPPM signal has a different contour other than db3, the method as described can be automated using several methods. As a particular signal can be usually represented by the same wavelet, this needs to be performed very few times during the model building.
[0040] In another embodiment of the invention, in the wavelet decomposition module (210), the PPGPPM is decomposed to extract wavelet features from the extracted photoplethysmogram (PPG) signals. Wavelet transform provides multi-resolution analysis and finer time localization is achieved at higher frequencies. Discrete Wavelet Transform (DWT) is applied on the PPGPPM signal to obtain the detail and approximation coefficients to transform the signal from its original form to a modified form, which is more suitable for further processing. The DWT decomposition step is performed numerous times on the approximation coefficients and is taken care of in the available DWT implementations, when the decomposition level is specified. The approximation coefficients and the detailed coefficients are collected from each level.
[0041] In another embodiment of the invention, in the feature computation module (212), the extracted features are computed, wherein, in an exemplary embodiment, the said extracted features can be standard deviation of cycle duration over a time window, relative crest time (Ts/Tc; Ts: Systolic Upstroke Time, Tc: Cycle Duration), relative diastolic time (Td/Tc; Td: Diastolic Time, Tc: Cycle Duration), time ratio (Ts/Td; Ts: Systolic Upstroke Time, Td: Diastolic Time), mean as derived from wavelet coefficients, variance as derived from wavelet coefficients, mean of energy as derived from wavelet coefficients, maximum amplitude as derived from wavelet coefficients, minimum amplitude as derived from wavelet coefficients, maximum energy as derived from wavelet coefficients, minimum energy as derived from wavelet coefficients, average frequency as derived from wavelet coefficients, mid frequency as derived from wavelet coefficients, maximum frequency as derived from wavelet coefficients, minimum frequency as derived from wavelet coefficients and half point of the energy as derived from wavelet coefficients.

[0042] In another embodiment of the invention, in the feature computation module (212), in an exemplary embodiment, the most relevant statistical and entropy related features are selected from a group comprising of standard deviation of cycle duration over a time window, relative crest time (Ts/Tc; Ts: Systolic Upstroke Time, Tc: Cycle Duration), relative diastolic time (Td/Tc; Td: Diastolic Time, Tc: Cycle Duration), time ratio (Ts/Td; Ts: Systolic Upstroke Time, Td: Diastolic Time), mean as derived from wavelet coefficients, variance as derived from wavelet coefficients, mean of energy as derived from wavelet coefficients, maximum amplitude as derived from wavelet coefficients, minimum amplitude as derived from wavelet coefficients, maximum energy as derived from wavelet coefficients, minimum energy as derived from wavelet coefficients, average frequency as derived from wavelet coefficients, mid frequency as derived from wavelet coefficients, maximum frequency as derived from wavelet coefficients, minimum frequency as derived from wavelet coefficients and half point of the energy as derived from wavelet coefficients.
[0043] In another embodiment of the invention, in the classification module (214), classification of coronary artery disease (CAD) patients and non-coronary artery disease (CAD) patients is performed by using the selected most relevant statistical and entropy features, wherein the classification of coronary artery disease (CAD) patients and non-coronary artery disease (CAD) patients is performed by using machine learning methods. Two different classification results are fused for a better accuracy. The classification module (214) performs the classification using time domain features and peak rate related features, followed by fusion to boost up overall confidence score.

I/We Claim:
1. A method for identifying coronary artery disease (CAD) patients using
photoplethysmogram (PPG) signal, said method comprising:
a. extracting photoplethysmogram (PPG) signals from a patient using an image
capturing device (202) coupled with a mobile communication device (204);
b. determination of PPG peak rate per minute (PPGPPM) from the extracted
photoplethysmogram (PPG) signals and extraction of time and frequency domain
features from the extracted photoplethysmogram (PPG) signals using a feature
extraction module (206);
c. selection of most matching mother wavelet using PPGPPM with respect to a
particular photoplethysmogram (PPG) signal window using a wavelet selection
module (208);
d. decomposition of the PPGPPM signal to extract wavelet features from the
extracted photoplethysmogram (PPG) signals using a wavelet decomposition
module (210);
e. computation of the extracted features and selection of the most relevant statistical
and entropy related features using a feature computation module (212);
f. classification of coronary artery disease (CAD) patients and non-coronary artery
disease (CAD) patients by using the selected most relevant statistical and entropy
features using a classification module (214);
2. The method as claimed in claim 1, wherein the photoplethysmogram (PPG) signals are extracted from a user’s peripheral body parts.
3. The method as claimed in claim 2, wherein the user’s peripheral body parts are selected from a group comprising of fingertip, ear, toe; and forehead.

4. The method as claimed in claim 1, wherein the photoplethysmogram (PPG) signals are extracted from a user using a light emitting source attached to the image capturing device (202) coupled with the mobile communication device (204).
5. The method as claimed in claim 1, wherein the image capturing device (202) coupled with the mobile communication device (204) extracts photoplethysmogram signals (PPG) as a video stream.
6. The method as claimed in claim 1, wherein the peak rate per minute (PPGPPM) is interpolated from number of peaks per second.
7. The method as claimed in claim 6, wherein said number of peaks per second is calculated by reciprocating each peak to next peak interval.
8. The method as claimed in claim 6, wherein the number of peaks per second value is obtained for each peak interval.
9. The method as claimed in claim 6, wherein the number of peaks per second value is resampled at uniform spacing and taking one sample per second.
10. The method as claimed in claim 7, wherein said peak intervals are spaced non-uniformly.
11. The method as claimed in claim 1, wherein the most matching mother wavelet is chosen from a set of mother wavelets and their correlation with the particular photoplethysmogram (PPG) signal window.
12. The method as claimed in claim 1, wherein said decomposition is stopped if the classification accuracy remains similar.
13. The method as claimed in claim 1, wherein the classification of coronary artery disease (CAD) patients and non-coronary artery disease (CAD) patients is performed by using machine learning methods.

14. A system (200) for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal, said system comprising:
a. an image capturing device (202) coupled with a mobile communication device
(204), adapted for extracting photoplethysmogram signals from a patient;
b. a feature extraction module (206) adapted for determination of PPG peak rate
per minute (PPGPPM) from the extracted photoplethysmogram (PPG) signals
and extraction of time and frequency domain features from the extracted
photoplethysmogram (PPG) signals ;
c. a wavelet selection module (208) adapted for selection of most matching
mother wavelet using PPGPPM with respect to a particular
photoplethysmogram (PPG) signal window;
d. a wavelet decomposition module (210) adapted for decomposition of the
PPGPPM signal to extract wavelet features from the extracted
photoplethysmogram (PPG) signals;
e. a feature computation module (212) adapted for computation of the extracted
features and selection of the most relevant statistical and entropy related
features;
f. a classification module (214) adapted for classification of coronary artery
disease (CAD) patients and non-coronary artery disease (CAD) patients by
using the selected most relevant statistical and entropy features.

ABSTRACT
METHOD AND SYSTEM FOR IDENTIFYING CORONARY ARTERY DISEASE PATIENTS USING PHOTOPLETHYSMOGRAM SIGNAL
A method and system for identifying coronary artery disease (CAD) patients using photoplethysmogram (PPG) signal is provided. The method is performed by extracting photoplethysmogram (PPG) signals from a patient, determining PPG peak rate per minute (PPGPPM) from the extracted photoplethysmogram (PPG) signals and extracting of time and frequency domain features from the extracted photoplethysmogram (PPG) signals, selecting most matching mother wavelet using PPGPPM with respect to a particular photoplethysmogram (PPG) signal window, decomposing the PPGPPM signal to extract wavelet features from the extracted photoplethysmogram (PPG) signals, computing the extracted features and selection of the most relevant statistical and entropy related features and classifying coronary artery disease (CAD) patients and non-coronary artery disease (CAD) patients by using the selected most relevant statistical and entropy features.