BriefCase-Triage: CARE Multi-Triage CT for Pneumothorax; Pericardial effusion; Large aortic aneurysm; Shoulder fracture or dislocation device

BriefCase-Triage: CARE is a radiological computer-aided triage and notification software that analyzes contrast and non-contrast CT images of the chest, abdomen, or both in adults aged 18+. The device uses artificial intelligence to flag suspected findings for pneumothorax, pericardial effusion, large aortic aneurysm, and shoulder fracture or dislocation, presenting compressed preview images to assist clinicians in workflow prioritization. The device does not alter original images, is not for diagnostic use beyond notification, and operates in parallel to standard care reading queues.

The subject device uses a multi-triage module with a locked AI algorithm fine-tuned from a foundation model, compared to the predicate's single-triage module. Both are deep learning algorithms integrated with image communication platforms and desktop applications. The subject device accepts both contrast and non-contrast CT images versus the predicate's contrast-only requirement. Both provide unannotated, compressed, low-quality, grayscale preview images and operate in parallel to standard workflow without removing or deprioritizing cases.

Not stated in this summary. The document mentions ISO 13485 clauses related to design controls, nonconforming product, and corrective/preventative action in regulatory requirements sections, but does not cite specific consensus standards for device testing.

The subject device is substantially equivalent to the predicate (BriefCase-Triage for AD, K251406) because both are AI-based radiological triage and notification software with identical intended use and operating principles: flagging time-sensitive findings, providing prioritization alerts with preview images for informational purposes only, not altering original images, and operating in parallel to standard care. Performance testing showed all four clinical indications exceeded pre-specified AUC and sensitivity/specificity goals (>95% and >80%), and time-to-notification results demonstrated comparable clinical benefit to the predicate.