Company Spotlight: Keya Medical DEEPVESSEL Plaque - Automated Coronary Plaque Phenotyping

We unpack the FDA 510(k) summary for Keya Medical's DEEPVESSEL Plaque, a cloud-based software for analyzing coronary CTA scans. The episode explores how the device uses deep learning and remote human analysts to quantify plaque volume, the performance metrics from its 147-patient validation study, and the operational shift from local workstation software to AI-as-a-service.

Key points

• DEEPVESSEL Plaque is cleared for patients aged 22 and older with suspected or known coronary artery disease, outputting total, calcified, non-calcified, and low-density plaque volumes.
• Unlike its predicate device which runs on local hospital workstations, this software operates via the cloud, with trained analysts in regional data centers reviewing and editing the AI's segmentation.
• In a 147-patient retrospective validation study, the device demonstrated a 0.976 Pearson correlation for total plaque volume when compared to a consensus of three expert readers.
• The validation dataset included diverse CT scanner types but had limited demographic representation for Black patients, who made up only 2.7% of the cohort.
• The FDA summary includes a specific warning that the software's use of interpolated data could mask pathology that is near or smaller than the scanning resolution.

Source: FDA 510(k) summary K260497: DEEPVESSEL Plaque (Keya Medical Technology) - U.S. Food and Drug Administration, 2026

This episode is an AI-generated conversation summarising a public document; the hosts' voices are synthetic. It is for information only and is not medical advice. Always refer to the original source.

Company: Keya Medical Technology (keyamedical.com) · Product: DEEPVESSEL Plaque · FDA 510(k) K260497

Full transcript: https://ai-in-medicine-podcast.vercel.app/episodes/company-spotlight-keya-medical-deepvessel-plaque-automated-coronary-pl-7a2b46