This white paper introduces HeartAssist™, an AI-powered solution designed to automate measurements in adult transthoracic echocardiography. Built on deep learning algorithms by ON:TACT HEALTH and integrated into Samsung Medison’s V7 and V8 ultrasound machines, HeartAssist™ enhances echocardiography by automating view classification and cardiac measurements, improving diagnostic consistency and efficiency.
In traditional transthoracic echocardiogram procedures, clinicians manually perform multiple measurements across cardiac cycles—a time-consuming and variable process. HeartAssist™ addresses this by fully automating the workflow: identifying views, selecting cardiac phases (end-diastole and end-systole), and conducting measurements across B-mode, M-mode, and Doppler imaging.
The system supports 20 echocardiographic views and delivers automated measurements with accuracy rates up to 98% for view classification and 95–100% for measurements, verified by clinical experts. This automation reduces manual keystrokes by up to 93% and measurement time by 90%, minimizing clinician workload and inter-observer variability.
Key capabilities include:
- Automated phase selection without requiring ECG, ensuring flexibility even when ECG signals are unavailable.
- Multi-cycle measurements, addressing inter-beat variability and aligning with clinical guidelines for accurate cardiac assessment.
- High accuracy in key parameters like left ventricular volume, wall thickness, and Doppler flow velocities.
By integrating AI in cardiology, HeartAssist™ transforms cardiac ultrasound workflows, providing a scalable solution for echocardiography labs. It enhances the precision of transthoracic echocardiogram procedures while reducing the physical demands on clinicians, contributing to more reliable and efficient cardiac diagnostics.
To explore how HeartAssist™ works in practice, check out the Quick Guide to HeartAssist™.
