Article | Apr 30, 2024
A study conducted at Christ Hospital Health Network and presented at ACC.24 showed good to excellent agreement between AI and experts assessments of LV volumes and ejection fraction in contrast echocardiograms.
Key Findings
- This is the first study to validate Us2.ai's fully automated LV assessment specifically on contrast echocardiograms, a common imaging modality used when non-contrast image quality is suboptimal, demonstrating the software's applicability beyond standard echo.
- AI-derived measurements of LV volumes and ejection fraction showed good to excellent agreement with expert human readers across contrast echocardiographic studies performed at Christ Hospital Health Network.
- The automated workflow required no manual interaction, processing contrast echo studies with the same pipeline used for non-contrast images, confirming vendor-neutral feasibility in a real-world clinical setting.
- Performance of the AI was comparable to that of experienced human readers, supporting its potential as a time-saving tool for LV assessment in patients who routinely receive contrast agents during echocardiography.
Palmer, C., Patel, K., Frost, M., Equilbec, C., Swaminathan, M., & Mazur, W. (2024). FULLY AUTOMATED ARTIFICIAL INTELLIGENCE ASSESSMENT OF THE LEFT VENTRICLE BY CONTRAST ECHOCARDIOGRAPHY. Journal of the American College of Cardiology, 83(13), 1424-1424. https://doi.org/10.1016/s0735-1097(24)03414-4