Deep Learning for Heart Failure Diagnostics and Therapeutic Discovery

Advancing Heart Failure Care With AI-Enhanced Diagnostics

Through the Asian neTwork for Translational Research and Cardiovascular Trials (ATTRaCT), A*STAR is pioneering new approaches in heart failure diagnostics and therapeutic discovery. By utilising deep learning and AI-driven tools like US2.AI, this initiative enhances echocardiography with improved speed, accuracy, and scalability. Spanning 50 sites across 12 countries, the ATTRaCT network is revolutionising how clinicians diagnose and manage heart failure, enabling faster clinical decisions and more personalised patient care.

The objectives

Enhancing Diagnostic Efficiency with AI

The primary goal of this initiative is to utilise AI and deep learning to streamline heart failure diagnostics. By reducing analysis time from 30 minutes to just 2 minutes, tools like US2.AI empower clinicians to make quicker, more accurate decisions, improving patient outcomes and reducing variability in diagnostics.

Facilitating Drug Discovery and Longitudinal Monitoring

The ATTRaCT network aims to support drug discovery and longitudinal surveillance for heart failure patients. By leveraging AI-driven research tools, the project provides insights into disease progression and therapeutic targets, ultimately advancing personalised treatment options and improving long-term care.

Scaling Data for Machine Learning Applications

A crucial objective is to build a scalable infrastructure capable of handling large datasets from 50 sites across 12 countries. By preparing DICOM data for machine learning, the project lays the foundation for extensive data analysis and predictive modeling, which are essential for the continuous development of AI-powered cardiovascular solutions.

The impact

The ATTRaCT network’s adoption of AI-based tools, such as US2.AI, has drastically accelerated heart failure diagnostics by reducing analysis time from 30 minutes to just 2 minutes. This improvement allows clinicians to make faster, more accurate decisions with minimal variability, ultimately enhancing patient care. In addition to streamlining diagnostics, the project has advanced cardiovascular drug discovery and patient monitoring. By leveraging deep learning, the initiative supports comprehensive data analysis that identifies therapeutic targets and enables long-term health tracking, paving the way for personalised treatment options.

The network’s global reach is also notable, spanning 50 sites across 12 countries and establishing a scalable platform for DICOM data collection. This vast data infrastructure supports large-scale machine learning applications, positioning A*STAR to make significant contributions to cardiovascular care on an international scale. Furthermore, US2.AI's mobile, non-invasive echocardiography decision support tool offers a cost-effective solution, expanding access to high-quality diagnostics, especially in areas with limited traditional resources. This approach broadens the accessibility of quality healthcare, making advanced cardiovascular care more inclusive and affordable.