Israel’s Aidoc artificial-intelligence platform reduced backlogs for urgent cases and prevented a cumulative delay of more than 1,500 hours in 2024, according to an Assuta Medical Center analysis of more than 5,000 computed tomography (CT) head scans.
The system also detected 21 cases of intracranial hemorrhages that were not flagged in the original radiology reports, enabling earlier diagnosis and potentially life-saving treatment, the study said.
Aidoc is an Israeli company whose triage tools are deployed widely in local hospitals.
Overall, the software achieved 89.9% sensitivity for detecting intracranial hemorrhage and 99% specificity. For prioritizing emergent studies in the clinical work list, it attained 100% sensitivity, meaning no true emergencies were missed. By comparison, imaging technicians who were asked to presort urgent cases had 66.7% sensitivity.
“This is a real-time safety net,” said Dr. Arnon Makori, head of imaging at Assuta Medical Centers, Israel’s largest private hospital system.
“The system alerts on urgent findings such as brain bleeds and lets the imaging specialist and radiologist decide immediately whether to send the patient to the emergency department,” he said. “For the first time in Israel, it has been implemented in an outpatient setting, not only in ERs, which is what makes it unique. It is not just a rapid-detection tool; it also shortens actual wait times for scans.”
Israel has a limited number of CT scanners, so any solution that streamlines reading scans improves patient care by accelerating radiologist review and decision-making, Makori said.
AI pivotal for triage as well as detection
Dr. Royi Barnea, the chief investigator at Assuta Health Services Research Institute, said the research underscored the value of AI for triage as well as detection.
“When there are many scans in the queue, it is not always easy for staff to decide who goes first,” he said. “Here, AI excels. It identifies emergencies with perfect accuracy and moves them to the front immediately. This is a significant addition to Israel’s healthcare system as it struggles with growing demand.”
Assuta plans to expand its use of AI to magnetic resonance imaging (MRI) and to screening pathways such as low-dose CT for lung cancer detection, Maariv reported.
Aidoc said its tools have shown high diagnostic accuracy for intracranial hemorrhage in multiple clinical evaluations, supporting their use to “watch” radiology work lists and escalate time-critical cases.