| CPC A61B 5/0205 (2013.01) [A61B 5/024 (2013.01); A61B 5/0816 (2013.01)] | 13 Claims |
|
1. A processor-implemented method for determining a breathing rate and a heart rate from cardiopulmonary signals, the method comprising the steps of:
receiving, via one or more hardware processors, a cardiopulmonary signal of a subject being monitored from a sensor unit among a sensor group including a pneumotachograph, an ultrasound sensor, an accelerometer including a wearable accelerometer, a respiratory belt, and a radar, wherein if the sensor unit is within a system, then the system is placed beside or near to the subject being monitored to acquire the cardiopulmonary signal, wherein if the sensor unit is connected to the system but present outside the system, then the sensor unit is placed beside or near to the subject being monitored to acquire the cardiopulmonary signal of a subject wearing the sensor unit, and wherein the cardiopulmonary signal is a continuous and an unobtrusive time-domain signal represented in seconds;
splitting, via the one or more hardware processors, the cardiopulmonary signal into one or more sub-cardiopulmonary signals, based on a predefined cardiopulmonary signal window duration, wherein the predefined cardiopulmonary signal window duration is at least 10 seconds or multiples of ‘5’ but not ‘5’, based on a domain knowledge;
pre-processing, via the one or more hardware processors, each sub-cardiopulmonary signal of the one or more sub-cardiopulmonary signals, to obtain a preprocessed sub-cardiopulmonary signal, for each sub-cardiopulmonary signal;
filtering, by a bandpass filter implemented via the one or more hardware processors connected or communicated with a portable or handheld device wirelessly or with a wired connection, through input/output (I/O) interface(s), the pre-processed sub-cardiopulmonary signal of each sub-cardiopulmonary signal, to obtain a first filtered signal and a second filtered signal, wherein the first filtered signal has frequencies between a predetermined first range of frequencies corresponding to breathing associated with the subject being monitored, and the second filtered signal has frequencies between a predetermined second range of frequencies corresponding to heartbeats associated with the subject being monitored, wherein the predetermined first range of frequencies include 0.2 Hz to 0.8 Hz, which is identified as general breathing frequency range, and wherein the predetermined second range of frequencies include 1 Hz to 2 Hz, which is identified as general heartbeats frequency range;
determining, via the one or more hardware processors, a breathing rate of the subject being monitored, from the first filtered signal, by using a spectrum analysis technique;
post-processing, via the one or more hardware processors, the second filtered signal to obtain a post-processed signal, wherein the post-processing comprises removing low frequency components underlying in the second filtered signal;
obtaining, via the one or more hardware processors, an overall signal quality value of each sub-cardiopulmonary signal, and a heart signal quality value of the second filtered signal corresponding to each sub-cardiopulmonary signal, wherein obtaining the overall signal quality value of each sub-cardiopulmonary signal comprises:
generating an overall signal autocorrelation curve whose envelope has a triangular contour, by applying an autocorrelation function on the sub-cardiopulmonary signal;
defining an area under the envelope of the overall signal autocorrelation curve as the overall signal quality value for the sub-cardiopulmonary signal; and
wherein obtaining the heart signal quality value of the second filtered signal corresponding to each sub-cardiopulmonary signal comprises:
generating a heart signal autocorrelation curve whose envelope has the triangular contour, by applying the autocorrelation function on the second filtered signal corresponding to the sub-cardiopulmonary signal; and
defining an area under the envelope of the heart signal autocorrelation curve as the heart signal quality value for the second filtered signal corresponding to the sub-cardiopulmonary signal;
determining, via the one or more hardware processors, a heart rate of the subject being monitored, using one of a set of signal processing techniques on the post-processed signal, based on the overall signal quality value and the heart signal quality value, wherein the set of signal processing techniques comprises (i) a sliding window based additive spectra technique when (i) the overall signal quality value is greater than a predefined overall signal quality upper threshold value, and (ii) the heart signal quality value is greater than a predefined heart signal quality upper threshold value, (ii) a mean peak to peak time difference technique when (i) the overall signal quality value is greater than a predefined overall signal quality lower threshold value, or (ii) (a) the overall signal quality value is greater than a predefined overall signal quality middle threshold value and (b) the heart signal quality value is greater than a predefined heart signal quality lower threshold value, and (iii) a minimum variance sweep technique when (i) the overall signal quality value is greater than a predefined overall signal quality middle threshold value, and (ii) the heart signal quality value is greater than a predefined heart signal quality upper threshold value, wherein the minimum variance sweep technique finds a signal duration where a variance of the mean peak to peak time difference is minimum such that a particular signal duration contain useful information related to the heart rate; and,
detecting vital signs including the breathing rate and the heart rate from the cardiopulmonary signals of the subject to be monitored, achieving long-term and effective health monitoring of healthy subjects, patients and infant subjects, and wherein the method is capable of being used in medical care units, hospitals, fitness tracking, stress monitoring, and presence detection in rescue operations.
|