Authors: Nathan T. Riek, Murat Akcakaya, Zeineb Bouzid, Tanmay Gokhale, Stephanie Helman, Karina Kraevsky-Philips, Rui Qi Ji, Ervin Sejdic, Jessica K. Zègre-Hemsey, Christian Martin-Gill, Clifton W. Callaway, Samir Saba, Salah Al-Zaiti

Published on: May 08, 2024

Impact Score: 7.4

Arxiv code: Arxiv:2405.09567

Summary

• What is new: Introduction of ECG-SMART-NET, a modified ResNet-18 architecture tailored to better capture temporal and spatial features in ECG data for OMI detection.
• Why this is important: Difficulty in visually identifying occlusion myocardial infarction (OMI) in ECGs, leading to potentially fatal delays in treatment.
• What the research proposes: A clinically informed modification of the ResNet-18 architecture that integrates temporal and spatial convolutional layers specifically designed for ECG data.
• Results: ECG-SMART-NET achieved a test AUC score of 0.889 and a test average precision score of 0.587, outperforming other models.

Technical Details

Technological frameworks used: Modified ResNet-18

Models used: Temporal and spatial convolutional layers

Data used: 10,893 ECGs from 7,297 unique patients

Potential Impact

Healthcare providers, medical device companies specializing in cardiac care, and ECG technology developers

Want to implement this idea in a business?

We have generated a startup concept here: HeartGuardAI.