HEART RATE DETECTION METHOD AND DEVICE USING HEART SOUND ACQUIRED FROM AUSCULATION POSITIONS

• US 20170071564A1
• Filed: 09/10/2015
• Published: 03/16/2017
• Est. Priority Date: 09/10/2015
• Status: Active Grant

• Current Assignee: Imediplus Inc.
• Original Assignee: Imediplus Inc.
• Inventors: WAN, LEI, CHEN, CHUNG LUN, LIAO, WEN LING, CHEN, YU HSUAN, TSAI, Kun-Hsi, YANG, Shih-I, KU, Shih-Hsuan, LIANG, Tzu-Chen
• Granted Patent: US 10,004,473 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: A61B 2505/01 Emergency care
• CPC Class Codes: A61B 5/0205 Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
• CPC Class Codes: A61B 5/02438 with portable devices, e.g. worn by the patient
• CPC Class Codes: A61B 5/0245 by using sensing means generating electric signals, i.e. ECG signals
• CPC Class Codes: A61B 5/0255 Recording instruments specially adapted therefor
• CPC Class Codes: A61B 5/316 Modalities, i.e. specific diagnostic methods
• CPC Class Codes: A61B 5/318 Heart-related electrical modalities, e.g. electrocardiography [ECG]
• CPC Class Codes: A61B 5/333 Recording apparatus specially adapted therefor
• CPC Class Codes: A61B 5/361 Detecting fibrillation
• CPC Class Codes: A61B 5/363 Detecting tachycardia or bradycardia
• CPC Class Codes: A61B 5/725 using specific filters therefor, e.g. Kalman or adaptive filters A61B5/316 take precedence
• CPC Class Codes: A61B 7/00 Instruments for auscultation
• CPC Class Codes: A61B 7/04 Electric stethoscopes microphones, acoustic transducers therefor H04R

Abstract

A heart rate detection method for calculating heart rate using heart sound from auscultation positions identified by a statistical approach utilizes a down-sampling and filtering process to acquire samples of heart sound from multiple auscultation positions of multiple testees and calculate heart rate with the samples, records time for calculating heart rate from each auscultation position of each testee and record the same from electrocardiogram, calculates a mean error and a standard deviation of the time to identify the auscultation positions allowing faster speed in heart rate detection, and applies a Bland-Altman difference plot and both a coefficient of determination and a Pearson'"'"'s correlation coefficient to determine the degree of consistency and correlation of the heart rate measured from the multiple auscultation positions to identify the auscultation positions allowing generation of precise heart rate.

Claims

1. A heart rate detection method using heart sounds acquired from auscultation positions, wherein the heart rate detection method is built in a heart rate detection device and is performed by a processor of the heart rate detection device, the heart rate detection method comprising steps of:

repeatedly acquiring and processing samples of heart sound at a first sampling frequency from multiple auscultation positions of multiple testees using a heart sound identification and heart rate detection process to identify a first heart sound occurring at beginning of each systole and a second heart sound occurring at beginning of a diastole paired to the systole from the samples acquired at a second sampling frequency, wherein the multiple auscultation positions include an auscultation position for mitral valve, an auscultation position for pulmonary valve, a first auscultation position for aortic valve, a second auscultation position for aortic valve, and an auscultation position for tricuspid valve;

calculating a target heart rate for each auscultation position of each testee according to each pair of the first heart sound and the second heart sound, and recording a target heart rate detection time for the auscultation position of the testee when calculation of the target heart rate for the auscultation position of the testee is completed for the first time; and

recording a reference heart rate for each auscultation position of each testee and a reference heart rate detection time for the auscultation position of the testee when calculation of the reference heart rate for the auscultation position is completed by a reference electrocardiogram (ECG) heart rate detector for the first time;

wherein the target heart rate detection time, the reference heart rate detection time, the target heart rate and the reference heart rate for each auscultation position are analyzed by a statistical process to obtain an analysis result that the target heart rate detection time measured from the auscultation position for tricuspid valve, the auscultation position for pulmonary valve and the auscultation position for mitral valve is faster than the target heart rate measured from any other auscultation positions, the target heart rate detection time measured from the auscultation position for mitral valve and the auscultation position for tricuspid valve has higher degree of stability than the target heart rate detection time measured from any other auscultation positions, and the target heart rate measured from the tricuspid position and the mitral position is more accurate than the target heart rate measured from any other auscultation positions.

2. The heart rate detection method as claimed in claim 1, wherein the statistical process includes steps of:

adding up the target heart rate detection time for each auscultation position of all the testees to calculate an arithmetic average value of the accumulated target heart rate detection time as a target heart rate detection speed for the auscultation position;

adding up a difference between the target heart rate detection speed and the reference heart rate detection speed for each auscultation position of all the testees to take an arithmetic average of the accumulated difference as a mean error for the auscultation position;

applying standard deviation to the mean error for each auscultation position to obtain a standard deviation error for the auscultation position representing a degree of stability of the target heart rate for the auscultation position of all the testees, wherein a lower value of the standard deviation error represents a higher degree of stability; and

assessing a degree of consistency and a degree of correlation between the target heart rate for each auscultation position of all the testees and the reference heart rate for the auscultation position of all the testees using a Bland-Altman difference plot and both a coefficient of determination and a Pearson'"'"'s correlation coefficient respectively, wherein a higher degree of correlation represents that the target heart rate for a corresponding auscultation position is more accurate.

3. The heart rate detection method as claimed in claim 1, wherein the heart sound identification and heart rate detection process includes steps of:

performing a down sampling processing on the samples of heart sound acquired from the multiple auscultation positions of the multiple testees to lower the first sampling frequency down to the second sampling frequency;

applying a bandpass filter and configuring two frequency thresholds to filter out noises falling out of frequency ranges of the first heart sound and the second heart sound from the samples of heart sound sampled at the second sampling frequency;

acquiring absolute value of energy of the filtered samples, taking an average energy value of all the positive samples, configuring an energy threshold from the average energy value, and setting the energy of the samples below the energy threshold to be zero;

applying a simple moving average filter to continuously redraw each multiple continuous samples identified in the step of acquiring absolute value of energy of the filtered samples with an average energy value of the multiple continuous samples to smoothen the filtered samples to generate the samples in the form of continuous peaks with and smooth contour; and

searching peaks corresponding to each pair of the first heart sound and the second heart sound from the smoothened samples to calculate a corresponding target heart rate.

4. The heart rate detection method as claimed in claim 2, wherein the heart sound identification and heart rate detection process includes steps of:

performing a down sampling processing on the samples of heart sound acquired from the multiple auscultation positions of the multiple testees to lower the first sampling frequency down to the second sampling frequency;

applying a bandpass filter and configuring two frequency thresholds to filter out noises falling out of frequency ranges of the first heart sound and the second heart sound from the samples of heart sound sampled at the second sampling frequency;

acquiring absolute value of energy of the filtered samples, taking an average energy value of all the positive samples, configuring an energy threshold from the average energy value, and setting the energy of the samples below the energy threshold to be zero;

applying a simple moving average filter to continuously redraw each multiple continuous samples identified in the step of acquiring absolute value of energy of the filtered samples with an average energy value of the multiple continuous samples to smoothen the filtered samples to generate the samples in the form of continuous peaks with and smooth contour; and

searching peaks corresponding to each pair of the first heart sound and the second heart sound from the smoothened samples to calculate a corresponding target heart rate.

5. The heart rate detection method as claimed in claim 3, wherein in the step of performing a down sampling processing on the samples of heart sound, the target heart rate for each auscultation position of each testee at each second is calculated by combining the samples of heart sound sampled in the second and the samples of heart sound sampled in previous two seconds ahead of the second.

6. The heart rate detection method as claimed in claim 4, wherein in the step of performing a down sampling processing on the samples of heart sound, the target heart rate for each auscultation position of each testee at each second is calculated by combining the samples of heart sound sampled in the second and the samples of heart sound sampled in previous two seconds ahead of the second.

7. The heart rate detection method as claimed in claim 1, applied to diagnosis of loss of pulse, ventricular arrhythmia, ventricular tachycardia and ventricular fibrillation.

8. The heart rate detection method as claimed in claim 2, applied to diagnosis of loss of pulse, ventricular arrhythmia, ventricular tachycardia and ventricular fibrillation.

9. The heart rate detection method as claimed in claim 3, applied to diagnosis of loss of pulse, ventricular arrhythmia, ventricular tachycardia and ventricular fibrillation.

10. The heart rate detection method as claimed in claim 4, applied to diagnosis of loss of pulse, ventricular arrhythmia, ventricular tachycardia and ventricular fibrillation.

11. The heart rate detection method as claimed in claim 5, applied to diagnosis of loss of pulse, ventricular arrhythmia, ventricular tachycardia and ventricular fibrillation.

12. The heart rate detection method as claimed in claim 6, applied to diagnosis of loss of pulse, ventricular arrhythmia, ventricular tachycardia and ventricular fibrillation.

13. A heart rate detection device connected to an analog input unit receiving heart sounds in the form of analog signal measured from five auscultation positions including an auscultation position for mitral valve, an auscultation position for pulmonary valve, a first auscultation position for aortic valve, a second auscultation position for aortic valve, and an auscultation position for tricuspid valve, the heart rate detection device comprising:

an analog-to-digital (A/D) converter adapted to connect to the analog input module and converting the heart sounds into digital data;

a processor connected to the A/D converter, and receiving the digital data transmitted from the A/D converter;

a memory module connected to the processor and storing the digital signals; and

a heart rate detection module connected to the processor, receiving the digital data transmitted from the processor, and performing the down-sampling process, a band-pass filtering process, a time sequence (TT) defined filtering process, a TT SMA filtering process and a peak-locating process to acquire heart rate of each auscultation position;

wherein the target heart rate detection time measured from the auscultation position for tricuspid valve, the auscultation position for pulmonary valve and the auscultation position for mitral valve is faster than the target heart rate measured from any other auscultation positions, the target heart rate detection time measured from the auscultation position for mitral valve and the auscultation position for tricuspid valve has higher degree of stability than the target heart rate detection time measured from any other auscultation positions, and the target heart rate measured from the tricuspid position and the mitral position is more accurate than the target heart rate measured from any other auscultation positions.