AI Platform 2.2 (AIP002)

AI Platform 2.2 is software that processes ultrasound images to detect, measure, and calculate medical parameters for patients with suspected disease. It provides real-time quality score feedback to assist healthcare professionals conducting echocardiography, abdominal, and lung ultrasound scans in adult patients. The device includes modules for cardiac measurements (LVEF, IVC diameter, myocardium wall thickness), lung analysis (A-lines and B-lines), and quality assessment across cardiac, lung, and abdominal views.

The modification extends the Quality AI module to assess two additional abdominal ultrasound views (upper quadrant and pelvic) beyond the predicate's cardiac and lung support. The core technology remains unchanged: deep convolutional neural networks for segmentation, landmark detection, and classification. All other features—intended use, AI architecture, data inputs/outputs, user workflow, risk profile, and labeling—remain identical to the predicate device (AI Platform 2.0, K240953).

IEC 62304:2006/AC:2015 (Medical device software lifecycle processes), FDA Content of Premarket Submissions for Device Software Functions guidance (June 2023), and FDA Technical performance assessment of quantitative imaging in radiological device premarket submissions guidance (June 2022). ISO 14971 was applied for risk management.

AI Platform 2.2 is substantially equivalent because it maintains the same indications for use, core AI technology, system design, data workflow, and safety/risk profile as the predicate. The sole modification—adding abdominal views to Quality AI—does not change intended use or introduce new hazard types; validation testing on 184 patients with 200 clips (29,371 frames) demonstrated equivalent performance (ICC=0.94-0.95). Risk controls were implemented and verified to maintain consistency with the predicate's residual risk levels.