Authors: Andrés Bell-Navas, Nourelhouda Groun, María Villalba-Orero, Enrique Lara-Pezzi, Jesús Garicano-Mena, Soledad Le Clainche

Published on: April 30, 2024

Impact Score: 8.2

Arxiv code: Arxiv:2404.19579

Summary

• What is new: First-time use of the Higher Order Dynamic Mode Decomposition (HODMD) algorithm in medical data augmentation and feature extraction, alongside an adaptation of the Vision Transformer (ViT) for heart disease recognition from echocardiography images.
• Why this is important: The increase in medical data and the high mortality rate from heart diseases necessitate the development of an efficient, early recognition system for cardiac pathologies.
• What the research proposes: A novel deep learning framework utilizing HODMD for data augmentation and feature extraction, and a customized Vision Transformer model for analyzing echocardiography images to detect heart diseases.
• Results: The system outperforms existing methods, including pretrained Convolutional Neural Networks, in predicting heart conditions from echocardiography video sequences.

Technical Details

Technological frameworks used: Higher Order Dynamic Mode Decomposition (HODMD), Vision Transformer (ViT)

Models used: Customized Vision Transformer (ViT) for image analysis

Data used: Database of annotated echocardiography video sequences

Potential Impact

Healthcare industry; medical imaging and diagnostics companies; AI-driven healthcare analytics firms.

Want to implement this idea in a business?

We have generated a startup concept here: HeartGuardAI.