FDA clears Israeli firm Aidoc's single AI model for 14 emergency CT findings

In late January 2026 the US Food and Drug Administration cleared a new triage system from Aidoc, the Israeli medical-imaging AI company founded by Elad Walach, Guy Reiner and Michael Braginsky. The product reads abdominal CT scans and alerts radiologists when it spots one of a wide range of acute, potentially life-threatening findings, so that the most urgent patients can be looked at first.

What made the clearance notable was the technology underneath it. Most clinical AI tools approved so far have been narrow: one algorithm, trained to detect one condition. Aidoc's system instead runs on CARE, the company's in-house foundation model, a large general-purpose model that is then adapted to specific medical tasks. The clearance added 11 new acute indications to three that had been cleared before, bringing them together in one workflow. The company describes it as the first FDA clearance of a double-digit number of acute indications powered by a single AI model.

According to figures reported by Calcalist and HLTH, the 11 newly cleared indications averaged 97 percent sensitivity and 98 percent specificity in the pivotal study reviewed by the FDA. Aidoc also says the unified approach produces roughly a tenth as many false alarms as leading single-condition tools, which matters in busy emergency departments where clinicians can be overwhelmed by alerts. These performance numbers come from the company and its submission rather than from an independent trial.

The system is delivered through aiOS, Aidoc's platform for running and monitoring AI inside hospitals, which the company says has already analysed more than 100 million patient cases. Aidoc has said it plans to extend its foundation model across CT and X-ray workflows over the next 18 months and is working on automated draft reports. The company has raised about $370 million to date.