Article | SEP 12, 2025
Presented at ASE 2025, this study evaluated the feasibility of using a handheld echocardiography device equipped with AI software (Us2.ai) to assess left ventricular diastolic dysfunction (LVDD) in preoperative patients. The study compared results from the handheld device with those from standard cart-based echocardiography.
Findings suggest that the AI-enabled handheld device can comprehensively detect LVDD in preoperative patients, performing comparably to traditional cart-based exams. These results support the potential use of AI-assisted handheld echocardiography as a practical and effective tool for preoperative cardiac assessment.
Key Findings:
- Handheld and cart-based echocardiography measurements showed good to excellent agreement (ICC 0.72–0.98; P < 0.001).
- Some handheld and cart-based echo measurements showed slight variation, but all remained within acceptable ranges of agreement.
- For both ASE and BSE grades of LVDD, there was statistically significant agreement between Us2.ai handheld and cart-based measurements (Kohen’s Kappa 0.98 and 0.93; P < 0.001).
For both ASE and BSE grades of LVDD, there was statistically significant agreement between Us2.ai handheld and cart-based measurements (Kohen’s Kappa 0.98 and 0.93; P < 0.001).
