Authors: Muhammad Uzair Zahid, Aysen Degerli, Fahad Sohrab, Serkan Kiranyaz, Moncef Gabbouj

Published on: February 09, 2024

Impact Score: 8.27

Arxiv code: Arxiv:2402.06530

Summary

• What is new: A novel method for early myocardial infarction detection using a one-class classification algorithm in echocardiography, incorporating a multi-modal approach and a composite kernel.
• Why this is important: The challenge of early myocardial infarction detection with limited echocardiography data availability.
• What the research proposes: A specialized framework using Multi-modal Subspace Support Vector Data Description, leveraging multi-view echocardiography and optimizing feature transformation.
• Results: Achieved a geometric mean of 71.24% in MI detection accuracy, indicating a substantial improvement in early diagnosis.

Technical Details

Technological frameworks used: Multi-modal Subspace Support Vector Data Description

Models used: One-class Classification

Data used: HMC-QU dataset with multiple echocardiography views

Potential Impact

Healthcare technology companies specializing in diagnostic imaging, particularly those focusing on cardiac care

Want to implement this idea in a business?

We have generated a startup concept here: HeartScan AI.