Artificial intelligence is emerging as a powerful tool in identifying transthyretin cardiac amyloidosis (ATTR-CM), a condition that remains substantially underdiagnosed despite therapeutic advances. In this nonrandomized clinical trial, the AI model ATTRACTnet screened ECG waveforms, echocardiographic parameters, demographics, and amyloidosis-related diagnosis codes to flag high-risk patients across a multisite health system. Among 1471 AI-positive individuals, 256 met strict eligibility criteria, with 50 undergoing confirmatory diagnostic testing—and nearly half (48%) receiving a definitive ATTR-CM diagnosis.
Notably, 88% of diagnosed patients initiated therapy within 3 months, resulting in a more than 2.8-fold increase in detection compared with historical controls. These findings suggest that AI-augmented pathways may meaningfully shorten diagnostic delays, though randomized trials…
