Post-market surveillance
Watching a tool after it is approved and deployed, to catch problems that only show up in real-world use. For clinical AI it means tracking whether a live tool is still accurate, still processing on time, and still being used as intended, because regulatory approval at launch does not guarantee ongoing performance. Increasingly treated as an obligation rather than best-effort. Closely related to post-deployment monitoring.
Eight months after go-live, a scanner software update changes how the CT angiography series is labeled, and the rule that routes it to the large vessel occlusion tool stops matching for that scanner. No error appears, the radiologists still read every scan, and the early alert that mobilized the thrombectomy team quietly stops arriving for every stroke patient scanned on that machine. Track processed studies and median scan-to-alert time per scanner each week, and name the person who is told when either one moves.
Terms like this come up in real clinical scenarios across the HelloAI courses: bite-sized modules with verifiable certificates. An account takes one minute, no password needed.
Sign in →