Prototyping an End-to-End Multi-Modal Tiny-CNN for Cardiovascular Sensor Patches

TL;DR

This paper presents a lightweight multi-modal deep learning model for cardiovascular sensor data classification, optimized for resource-constrained edge devices, achieving significant reductions in memory and energy consumption while maintaining accuracy.

Contribution

The authors develop a novel end-to-end tiny-CNN with early data fusion for ECG and PCG classification, demonstrating its efficiency and applicability on medical edge hardware.

Findings

Model reduces memory and compute by three orders of magnitude.

On-device inference is more energy-efficient than continuous streaming.

The approach maintains competitive accuracy on Physionet dataset.

Abstract

The vast majority of cardiovascular diseases may be preventable if early signs and risk factors are detected. Cardiovascular monitoring with body-worn sensor devices like sensor patches allows for the detection of such signs while preserving the freedom and comfort of patients. However, the analysis of the sensor data must be robust, reliable, efficient, and highly accurate. Deep learning methods can automate data interpretation, reducing the workload of clinicians. In this work, we analyze the feasibility of applying deep learning models to the classification of synchronized electrocardiogram (ECG) and phonocardiogram (PCG) recordings on resource-constrained medical edge devices. We propose a convolutional neural network with early fusion of data to solve a binary classification problem. We train and validate our model on the synchronized ECG and PCG recordings from the Physionet Challenge 2016 dataset. Our approach reduces memory footprint and compute cost by three orders of magnitude compared to the state-of-the-art while maintaining competitive accuracy. We demonstrate the applicability of our proposed model on medical edge devices by analyzing energy consumption on a microcontroller and an experimental sensor device setup, confirming that on-device inference can be more energy-efficient than continuous data streaming.