BriefCase-Triage

BriefCase-Triage is a radiological computer-aided triage and notification software that analyzes CT images containing the ribs in adults aged 18 and older. It uses artificial intelligence to detect suspected cases of three or more acute rib fractures and alerts clinicians via pop-up notifications with low-quality preview images, enabling faster triage without altering original images or replacing standard diagnostic review.

The subject device is substantially identical to the predicate (K230020) in architecture: both use deep learning AI algorithms integrated with DICOM-compliant systems, both operate as parallel notification-only tools that do not remove cases from standard workflow queues, and both display unannotated compressed grayscale preview images for informational purposes. The main difference is the narrowed anatomical focus (ribs only versus full chest) and improvements in algorithm training process and time-to-notification performance (41.4 seconds versus 70.1 seconds for the predicate).

Not stated in this summary.

Both devices raise identical safety and effectiveness questions regarding accurate triage of findings while operating in parallel to standard care without disrupting workflow. The subject device achieves superior or equivalent performance metrics: AUC of 97.2% (predicate also met >0.95), sensitivity 95.2%, specificity 95.1%, and faster time-to-notification. Both use the same fundamental AI/deep-learning approach with DICOM input, neither removes or de-prioritizes cases, and both provide only notification alerts with non-diagnostic preview images, ensuring clinicians retain full diagnostic authority through standard PACS review.