Model-driven Heart Rate Estimation and Heart Murmur Detection based on Phonocardiogram

TL;DR

The paper addresses noninvasive heart rate estimation and heart murmur detection from phonocardiograms (PCGs) in real-world noisy settings. It compares CNN-based HR estimators and develops a multi-task 2dCNN-MTL model that jointly performs HR estimation and murmur detection on CirCor DigiScope data, using diverse acoustic features and a sliding-window approach. The key results show a best HR mean absolute error of $MAE=1.312$ bpm for the single-task HR model and murmur accuracy exceeding $97 ext{%}$ with the multi-task model, while maintaining $MAE_{HR}$ around $1.3$–$1.6$ bpm and satisfying AAMI guidelines. These findings support accurate, resource-efficient remote monitoring and self-care for cardiovascular health.

Abstract

Acoustic signals are crucial for health monitoring, particularly heart sounds which provide essential data like heart rate and detect cardiac anomalies such as murmurs. This study utilizes a publicly available phonocardiogram (PCG) dataset to estimate heart rate using model-driven methods and extends the best-performing model to a multi-task learning (MTL) framework for simultaneous heart rate estimation and murmur detection. Heart rate estimates are derived using a sliding window technique on heart sound snippets, analyzed with a combination of acoustic features (Mel spectrogram, cepstral coefficients, power spectral density, root mean square energy). Our findings indicate that a 2D convolutional neural network (\textbf{\texttt{2dCNN}}) is most effective for heart rate estimation, achieving a mean absolute error (MAE) of 1.312 bpm. We systematically investigate the impact of different feature combinations and find that utilizing all four features yields the best results. The MTL model (\textbf{\texttt{2dCNN-MTL}}) achieves accuracy over 95% in murmur detection, surpassing existing models, while maintaining an MAE of 1.636 bpm in heart rate estimation, satisfying the requirements stated by Association for the Advancement of Medical Instrumentation (AAMI).

Figures (10)

• Figure 1: The overall goal of this project.
• Figure 2: The training data preparation process.
• Figure 3: HR estimation based on time convolutional (1D) neural network and LSTM (TCNN-lstm).
• Figure 4: HR estimation based on 2D convolutional neural network (2dCNN).
• Figure 5: HR estimation based on a fusion 2D-convolutional neural network (2dCNN-Fusion).