We claim:
1. A system 100 for determining breathing rate as a biofeedback during a meditation session or an exercise session, the system 100 comprising:
a computation engine 122, 204 executed by a processor 124, 206 and configured to:
extract a first biomarker, a second biomarker, a third biomarker and a fourth biomarker from physiological parameters associated with a subject by applying a pre-defined set of rules, wherein the physiological parameters are received from a contactless sensor device 102; and
a feedback unit 128, 222 executed by the processor 124, 206 and configured to:
compute a first value in real-time as a function of the second biomarker, the third biomarker and the fourth biomarker;
determine a correlation between the first value and a time domain parameter of the fourth biomarker and a frequency domain parameter of the fourth biomarker, wherein the first value is indicative of a stress level of the subject;
compute a second value by maximizing the time domain parameter of the fourth biomarker and minimizing the frequency domain parameter of the fourth biomarker based on the correlation, wherein the second value is indicative of a reduced stress level of the subject; and
47

transmit a biofeedback, in real-time, to a cue generation unit 130, wherein the biofeedback is representative of a quantified data that is determined based on the second value, wherein the quantified data is indicative of a modified second biomarker.
The system 100 as claimed in claim 1, wherein the computation engine 122, 204 extracts the first biomarker from the physiological parameters associated with the subject by applying a first set of rules from the pre-defined set of rules; derive the second biomarker and the third biomarker from a selected element of the first biomarker, wherein the second biomarker and the third biomarker are derived by applying a second set of rules and a third set of rules respectively from the pre¬defined set of rules; and compute the fourth biomarker based on the third biomarker.
The system 100 as claimed in claim 1, wherein the computation engine 122, 204 receives the physiological parameters from a database 118 in a predetermined format.
The system 100 as claimed in claim 1, wherein the system 100 comprises a data capturing engine 106 executed by a processor 114 and configured to receive the physiological parameters from the contactless sensor device 102 and transmit the physiological parameters to the database 118.
The system 100 as claimed in claim 4, wherein the data capturing engine 106 comprises a data acquisition unit 108 executed by the processor 114 and configured to record micro-voltage digital signal corresponding to the physiological parameters, captured in the form of micro-vibrations by the contactless sensor device 102, in a chronological order.
48

6. The system 100 as claimed in claim 5, wherein the data capturing engine 106 comprises a conditioning unit 110 executed by the processor 114 and configured to amplify the micro-voltage digital signal for maximizing resolution of the micro-voltage digital signal, the maximizing includes amplification between the range of 15x to 2500x.
7. The system 100 as claimed in claim 1, wherein the computation engine 122, 204 comprises a body movement detection unit 212 executed by the processor 124, 206 and configured to apply the first set of rules on the physiological parameters to obtain the first biomarker, wherein the first biomarker represents a 'body movements' element, a 'no body movements' element and a 'body artifacts' element associated with the subject.
8. The system 100 as claimed in claim 7, wherein the first set of rules include processing the physiological parameters as multiple dataset points and applying a density based spatial clustering of applications with noise (DBSCAN) technique for clustering similar dataset points to identify the first biomarker and selecting the
'no body movement' element.
9. The system 100 as claimed in claim 8, wherein the computation engine 122, 204 comprises a first filtering unit 210 executed by the processor 124, 206 and configured to apply the second set of rules on the selected 'no body movement' element for deriving the second biomarker, the second biomarker represents respiratory signal in the form of a sinusoidal wave, and transmitting the second biomarker to a respiration extraction unit 214 within the computation engine 122, 204.
10. The system 100 as claimed in claim 9, wherein the respiration extraction unit 214 executed by the processor
49

124, 206 and configured to process the second biomarker for computing total number of maximas and minimas present in the sinusoidal wave.
11. The system 100 as claimed in claim 10, wherein the maximas and minimas of the sinusoidal wave are extracted separately into multiple templates.
12. The system 100 as claimed in claim 11, wherein the respiration extraction unit 214 executed by the processor 124, 206 and configured to apply K-means++ clustering for segregating the multiple templates into three template clusters.
13. The system 100 as claimed in claim 12, wherein the respiration extraction unit 214 is configured to select a first principal template from the three template clusters based on a Euclidean distance of the cluster centers, wherein the first principal template is representative of maximum number of respiration cycles.
14. The system 100 as claimed in claim 8, wherein the computation engine 122, 204 comprises a second filtering unit 220 executed by the processor 124, 206 and configured to apply the third set of rules on the selected 'no body movement' element for deriving the third biomarker that represents heartbeat signal in the form of multiple waveforms, and transmit the third biomarker to a heartbeat extraction unit 216 within the computation engine 122, 204.
15. The system 100 as claimed in claim 14, wherein the heartbeat extraction unit 216 is executed by the processor 124, 206 and configured to, process the third biomarker to form multiple heartbeat signal waveform templates, wherein signal between three continuous maximas and two continuous minimas of each of the
50

multiple heartbeat signal waveforms are processed to form the heartbeat waveform signal templates.
16. The system 100 as claimed in claim 15, wherein the heartbeat extraction unit 216 analyzes each of the multiple heartbeat signal waveform templates to assess similarities therebetween for clustering similar templates, wherein the heartbeat waveform signal templates are clustered into eight template clusters based on a frequency composition of the centroid template, frequency analysis technique, and Fast Fourier Transform (FFT) technique.
17. The system 100 as claimed in claim 16, wherein the heartbeat extraction unit 216 is configured to select a second principal template from the eight heartbeat signal waveform template clusters, wherein the second principal template is representative of maximum number of heartbeats.
18. The system 100 as claimed in claim 16, wherein the heartbeat extraction unit 216 is configured to analyze the clustered heartbeat waveform signal templates by based on a Pearson correlation technique to identify missing heartbeats by determining abnormal intervals between neighboring heartbeats.
19. The system 100 as claimed in claim 17, wherein the computation engine 122, 204 comprises a heart rate variability (HRV) detection unit 218 executed by the processor 124, 206 and configured to compute the fourth biomarker from the second principal template associated with the third biomarker, the fourth biomarker represents heart rate variability (HRV) parameters associated with the heartbeat signal, wherein the HRV parameters are determined based on time domain parameters and frequency domain parameters associated with each heartbeat.
51

20. The system 100 as claimed in claim 19, wherein the time domain HRV parameters includes standard deviation of normal-to-normal intervals (SDNN), standard deviation of the average normal-to-normal intervals (SDANN), root mean square of successive differences (RMSSD) and proportion of NN50 (pNN50); and frequency domain HRV parameters includes Very Low Frequency (VLF) , Low Frequency (LF) , High Frequency (HF) and ratio of LF to HF (LF/HF).
21. The system 100 as claimed in claim 1, wherein the first value is computed as a non-linear function of the second biomarker, the third biomarker and the fourth biomarker.
22. The system 100 as claimed in claim 1, wherein the correlation is representative of an inverse relationship between the first value and the time domain parameter of the fourth biomarker and a direct relationship between the first value and the frequency domain parameter of the fourth biomarker.
23. The system 100 as claimed in claim 1, wherein the feedback unit 128, 222 is configured to compute the second value by maximizing the time domain SDNN parameter and minimizing the frequency domain LF/HF parameter associated with the HRV parameters based on a regression model with back propagation.
24. The system 100 as claimed in claim 1, wherein the cue generation unit 130 is configured to provide the modified breathing rate in the form of at least an audio, a video and a haptic feedback.
25. The system 100 as claimed in claim 1, wherein the contactless sensor device 102 is configured to capture physiological parameters as a consequence of the modified second biomarker, transmit the physiological parameters to the computation engine 122, 204 via the data capturing engine 106, wherein the computation engine 122, 204
52

computes and transmits another set of first, second, third and fourth biomarkers to the feedback unit 128, 222 for computing a new second value and transmit another modified second biomarker.
26. A method for determining breathing rate as a biofeedback during a meditation session or an exercise session, the method comprising:
extracting, by a processor 124, 206, a first biomarker, a second biomarker, a third biomarker and a fourth biomarker from physiological parameters associated with a subject by applying a pre-defined set of rules, wherein the physiological parameters are received from a contactless sensor device 102;
computing, by the processor 124, 206, a first value in real-time as a function of the second biomarker, the third biomarker and the fourth biomarker;
determining, by the processor 124, 206, a correlation between the first value and a time domain parameter of the fourth biomarker and a frequency domain parameter of the fourth biomarker, wherein the first value is indicative of a stress level of the subject;
computing, by the processor 124, 206, a second value by maximizing the time domain parameter of the fourth biomarker and minimizing the frequency domain parameter of the fourth biomarker based on the correlation, wherein the second value is indicative of a reduced stress level of the subject; and
transmitting, by the processor 124, 206, a biofeedback, in real-time, to a cue generation unit 130, wherein the biofeedback is representative of a quantified data that is determined based on the second value, wherein the
53

quantified data is indicative of a modified second biomarker.
27. The method as claimed in claim 26, wherein the first biomarker is extracted from physiological parameters associated with the subject by applying a first set of rules from the pre-defined set of rules; the second biomarker and the third biomarker are derived from a selected element of the first biomarker, wherein the second biomarker and the third biomarker are derived by applying a second set of rules and a third set of rules respectively from the pre-defined set of rules; and the fourth biomarker is computed based on the third biomarker.
28. The method as claimed in claim 26, wherein the physiological parameters are received in a pre-determined format.
29. The method as claimed in claim 26, wherein micro-vibrations corresponding to the physiological parameters are recorded in the form of micro-voltage digital signal in a chronological order.
30. The method as claimed in claim 29, wherein the micro-voltage digital signal is amplified for maximizing resolution of the micro-voltage digital signal, the maximizing includes amplification between the range of 15x to 2500x.
31. The method as claimed in claim 26, wherein the first set of rules are applied on the physiological parameters to obtain the first biomarker, wherein the first biomarker represents a 'body movements' element, a 'no body movements' element and a 'body artifacts' element associated with the subject.
54

32. The method as claimed in claim 31, wherein the first set of rules include processing the physiological parameters as multiple dataset points and applying a density based spatial clustering of applications with noise (DBSCAN) technique for clustering similar dataset points to identify the first biomarker and selecting the 'no body movement' element.
33. The method as claimed in claim 32, wherein the second set of rules are applied on the selected 'no body movement' element for deriving the second biomarker, the second biomarker represents respiratory signal in the form of a sinusoidal wave and wherein the second biomarker is transmitted to the processor 124, 206.
34. The method as claimed in claim 33, wherein the second biomarker is processed for computing total number of maximas and minimas present in the sinusoidal wave.
35. The method as claimed in claim 34, wherein the maximas and minimas of the sinusoidal wave are extracted separately into multiple templates.
36. The method as claimed in claim 35, wherein K-means++ clustering is applied for segregating the multiple templates into three template clusters.
37. The method as claimed in claim 36, wherein a first principal template is selected from the three template clusters based on a Euclidean distance of the cluster centres, wherein the first principal template is representative of maximum number of respiration cycles.
38. The method as claimed in claim 32, wherein the third set of rules are applied on the selected 'no body movement' element for deriving the third biomarker, the third biomarker represents heartbeat signal in the form of
55

multiple waveforms, and wherein the third biomarker is transmitted to the processor 124, 206.
39. The method as claimed in claim 38, wherein the third biomarker is processed to form multiple heartbeat signal waveform templates, and wherein signal between three continuous maximas and two continuous minimas of each of the multiple heartbeat signal waveforms are processed to form the heartbeat waveform signal templates.
40. The method as claimed in claim 39, wherein each of the multiple heartbeat signal waveform templates are analyzed for assessing similarities therebetween for clustering similar templates, wherein the heartbeat waveform signal templates are clustered into eight template clusters based on a frequency composition of the centroid template, frequency analysis technique, and Fast Fourier Transform (FFT) technique.
41. The method as claimed in claim 40, wherein a second principal template is selected from the eight heartbeat signal waveform template clusters, wherein the second principal template is representative of maximum number of heartbeats.
42. The method as claimed in claim 40, wherein the clustered heartbeat waveform signal templates are analyzed based on a Pearson correlation technique to identify missing heartbeats by determining abnormal intervals between neighboring heartbeats.
43. The method as claimed in claim 41, wherein the fourth biomarker is computed from the second principal template associated with the third biomarker, the fourth biomarker represents heart rate variability (HRV) parameters associated with the heartbeat signal, wherein the HRV parameters are determined based on time domain parameters
56

and frequency domain parameters associated with each heartbeat.
44. The method as claimed in claim 43, wherein the time domain HRV parameters includes standard deviation of normal-to-normal intervals (SDNN), standard deviation of the average normal-to-normal intervals (SDANN), root mean square of successive differences (RMSSD) and proportion of NN50 (pNN50) and frequency domain HRV parameters includes Very Low Frequency (VLF), Low Frequency (LF) , High Frequency (HF) and ratio of LF to HF (LF/HF) .
45. The method as claimed in claim 26, wherein the first value is computed as a non-linear function of the second biomarker, the third biomarker and the fourth biomarker.
46. The method as claimed in claim 26, wherein the correlation is representative of an inverse relationship between the first value and the time domain parameter of the fourth biomarker and a direct relationship between the first value and the frequency domain parameter of the fourth biomarker.
47. The method as claimed in claim 26, wherein the second value is computed by maximizing the time domain SDNN parameter and minimizing the frequency domain LF/HF parameter associated with the HRV parameters based on a regression model with back propagation.
48. The method as claimed in claim 26, wherein the cue is provided in the form of at least an audio, a video and a haptic feedback.
49. The method as claimed in claim 26, wherein physiological parameters are captured as a consequence of the modified second biomarker, wherein another set of first, second, third and fourth biomarkers is computed and transmitted for deriving the new second value, and another modified
57

second biomarker is transmitted to the cue generation unit 130.
50. The method as claimed in claim 26, wherein a score of the meditation session or exercise session is generated based on at least; number of movements, change in heart rate, correlation of breathing rate with the breathing cue, coherence in breathing, change in breathing rate, change in first value computed based on HRV parameters, total time of meditation to the actual time of meditation and feedback from the subject.
Dated this 27th day of August, 2019.
Turtile Shell Technologies Private Limited
(Dev Robinson)
of Shardul Amarchand Mangaldas & Co.
Attorneys for the Applicant
58