BriefCase

Aidoc's BriefCase is an AI-powered radiological triage software that enables appropriately trained medical specialists to receive timely notifications for suspected critical findings across a suite of CT imaging applications including cervical spine, head, chest, and abdomen. The system uses an artificial intelligence algorithm to analyze DICOM images and provide non-diagnostic preview alerts to prioritize cases within the existing standard of care, aiming to reduce delays in attention to urgent conditions like fractures, hemorrhages, embolisms, and occlusions.

Aidoc's BriefCase is a radiological computer-aided triage and notification software designed to assist appropriately trained medical specialists in improving workflow efficiency. It analyzes various types of CT images—including cervical spine, non-enhanced head CT, CTPA, abdominal CT, contrast-enhanced chest CT (for incidental PE), and head CTA—to flag suspected positive findings for a range of pathologies. These pathologies include cervical spine fractures (linear lucencies), Intracranial Hemorrhage (ICH), Pulmonary Embolism (PE), Intra-abdominal Free Gas (IFG), incidental Pulmonary Embolism (iPE), and Large Vessel Occlusion (LVO). The software employs an artificial intelligence algorithm to process DICOM images, operating in parallel with standard image interpretation protocols. Its primary function is to expedite the triage process by alerting users to potentially critical cases, with a key update being the expansion of its indicated user base. The BriefCase system consists of server components (Aidoc Hospital Server and Aidoc Cloud Server) and a Worklist Application installed on the user's desktop. When the AI algorithm detects a suspected finding, it sends a notification to the Worklist Application, which includes patient information, the suspected pathology, and a compressed, unmarked preview image. These preview images are for informational purposes only and explicitly labeled "not for diagnostic use"; they do not mark findings or alter the original medical images. Clinicians are responsible for reviewing the full, original images per standard of care. This version of the device notably broadens the intended user population for these established functionalities from radiologists to "appropriately trained medical specialists," relying on previously established performance data for the underlying algorithms for each pathology.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. What are the sensitivity and specificity metrics of the AI algorithm for each of the six indicated pathologies (cervical spine fracture, ICH, PE, IFG, iPE, LVO) as established in the data supporting their initial functionalities?; How does the 'appropriately trained medical specialist' user category impact the interpretation workflow and time-to-diagnosis compared to when the tool was indicated only for radiologists?; What mechanisms are in place to ensure that the non-diagnostic preview images are not mistakenly used for diagnostic decisions, especially by the broadened user base?; Given that new performance data was not submitted for the broadened user indication, what was the rationale or evidence supporting the safety and efficacy of expanding use to non-radiologist specialists for these diverse and critical findings?; Are there any specific limitations or differences in how BriefCase should be used by different types of 'appropriately trained medical specialists' (e.g., emergency physicians vs. neurologists vs. trauma surgeons) for the various pathologies?