Authors: Md Jobayer, Md. Mehedi Hasan Shawon, Md Rakibul Hasan, Shreya Ghosh, Tom Gedeon, Md Zakir Hossain

Published on: May 10, 2024

Impact Score: 8.2

Arxiv code: Arxiv:2405.09570

Summary

• What is new: The paper introduces a novel end-to-end real-time heart murmur detection model using traditional and depthwise separable convolutional networks, designed to function on resource-constrained devices.
• Why this is important: Existing methods for detecting heart murmurs are limited by the need for extensive training, cost, accessibility, and susceptibility to noise interference.
• What the research proposes: A deep learning framework that uses continuous wavelet transform for feature extraction followed by a network consisting of Squeeze net, Bottleneck, and Expansion net to efficiently process PCG data.
• Results: Achieved state-of-the-art performance on four public datasets and demonstrated functionality on resource-constrained devices with up to 99.70% accuracy.

Technical Details

Technological frameworks used: Deep learning

Models used: Traditional and depthwise separable convolutional networks

Data used: Four publicly available PCG datasets

Potential Impact

Healthcare providers, medical device companies, and developers of diagnostic technologies might be impacted by the deployment of this cost-effective, highly accurate system.

Want to implement this idea in a business?

We have generated a startup concept here: HeartWise.