Deep Neural Networks for the Recognition and Classification of Heart Murmurs Using Neuromorphic Auditory Sensors
Insights
A novel AI tool using a neuromorphic auditory sensor and convolutional neural networks accurately classifies heart murmurs, improving diagnostic accuracy for cardiovascular diseases and reducing medical errors.
Area of Science:
- Cardiology and Artificial Intelligence
- Biomedical Signal Processing
- Machine Learning for Healthcare
Background:
- Auscultation is crucial for detecting cardiovascular diseases, but physician accuracy in identifying heart murmurs is insufficient, leading to diagnostic errors.
- Heart murmurs, common findings during auscultation, can be innocent or indicate serious heart conditions.
- Existing diagnostic methods have limitations in accuracy, contributing to Type-I and Type-II errors in clinical practice.
Purpose of the Study:
- To develop and evaluate a novel AI-powered tool for classifying heart sounds to differentiate between healthy individuals and pathological patients.
- To leverage a neuromorphic auditory sensor and convolutional neural networks (CNNs) for real-time audio decomposition and analysis of heart murmurs.
- To enhance the decision-making process for physicians during cardiac auscultation and minimize diagnostic errors.
Main Methods:
- Utilized a neuromorphic auditory sensor for FPGA to decompose heart sound recordings into frequency bands in real time.
- Generated sonogram images from the preprocessed audio data.
- Trained and tested various CNN architectures, including a modified AlexNet model, using heart murmur datasets from multiple research groups.
Main Results:
- The modified AlexNet model achieved a high accuracy of 97% in classifying heart sounds.
- Demonstrated strong performance with a specificity of 95.12% and sensitivity of 93.20%.
- Achieved a PhysioNet/CinC Challenge 2016 score of 0.9416, indicating robust diagnostic capability.
Conclusions:
- The developed AI tool shows significant potential in aiding physicians to accurately diagnose cardiovascular conditions through heart sound auscultation.
- The system's high accuracy and ability to reduce Type-I and Type-II errors can improve patient outcomes.
- This technology offers a promising advancement in non-invasive cardiac diagnostics, supporting primary care and specialist decision-making.
Abstract:
Auscultation is one of the most used techniques for detecting cardiovascular diseases, which is one of the main causes of death in the world. Heart murmurs are the most common abnormal finding when a patient visits the physician for auscultation. These heart sounds can either be innocent, which are harmless, or abnormal, which may be a sign of a more serious heart condition. However, the accuracy rate of primary care physicians and expert cardiologists when auscultating is not good enough to avoid most of both type-I (healthy patients are sent for echocardiogram) and type-II (pathological patients are sent home without medication or treatment) errors made. In this paper, the authors present a novel convolutional neural network based tool for classifying between healthy people and pathological patients using a neuromorphic auditory sensor for FPGA that is able to decompose the audio into frequency bands in real time. For this purpose, different networks have been trained with the heart murmur information contained in heart sound recordings obtained from nine different heart sound databases sourced from multiple research groups. These samples are segmented and preprocessed using the neuromorphic auditory sensor to decompose their audio information into frequency bands and, after that, sonogram images with the same size are generated. These images have been used to train and test different convolutional neural network architectures. The best results have been obtained with a modified version of the AlexNet model, achieving 97% accuracy (specificity: 95.12%, sensitivity: 93.20%, PhysioNet/CinC Challenge 2016 score: 0.9416). This tool could aid cardiologists and primary care physicians in the auscultation process, improving the decision making task and reducing type-I and type-II errors.
