Classifier ensemble for detection of abnormal heart sounds

Abstract

Various embodiments of the inventions of the present disclosure provide a combination of feature-based approach and deep learning approach for distinguishing between normal heart sounds and abnormal heart sounds. A feature-based classifier (60) is applied to a phonocardiogram (PCG) signal to obtain a feature-based abnormality classification of the heart sounds represented by the PCG signal and a deep learning classifier (70) is also applied to the PCG signal to obtain a deep learning abnormality classification of the heart sounds represented by the PCG signal. A final decision analyzer (80) isapplied to the feature-based abnormality classification and the deep learning abnormality classification of the heart sounds represented by the PCG signal to determine a final abnormality classification decision of the PCG signal.

Description

technical field [0001] Various embodiments described in this disclosure relate to systems, devices, and methods for detecting abnormal heart sounds. Background technique [0002] Cardiovascular disease (CVD) is a leading cause of morbidity and mortality worldwide, accounting for an estimated 17.5 million deaths in 2012. Cardiac auscultation is the main tool for screening and diagnosing CVD in primary health care. The use of digital stethoscopes and mobile devices offers clinicians the opportunity to record and analyze heart sounds (phonocardiogram, PCG) for diagnostic purposes. [0003] However, analysis of heart sound recordings in both clinical and non-clinical settings has proven to be a challenging task due to environmental noise (eg, alarms, talking). In addition, recording heart sounds by non-experts also increases the challenge for automated heart sound analysis. For example, changing the microphone position may alter the heart sound amplitude and may make the hear...

AI Technical Summary

Problems solved by technology

[0003] However, analysis of heart sound recordings in both clinical and non-clinical settings has proven to be challenging due to environmental noise (e.g., alarms, talking)

In addition, recording heart sounds by non-experts also adds challenges to automated heart sound analysis

For example, changing microphone position may alter heart sound amplitude and may make heart sounds prone to noise

Also, when heart sounds are recorded by different instruments, the quality of the heart sound recordings may vary (e.g. due to differences in the filter specifications of the different instruments), making it challenging to use a single algorithm

Due to the above factors, feature-based methods (traditional heart sound analysis) may have relatively low accuracy in the classification of abnormal heart sounds

Images