Thursday 30 November | 9:40-9:50 | R3-SSNR15-2 | Room S402
In this lecture, researchers from Tel Aviv University in Israel will describe how an AI algorithm can retrospectively detect intracranial aneurysms that had been missed by CT angiography (CTA) examinations.
Identifying intracranial aneurysms early is critical, as early detection facilitates risk stratification and timely, optimal management, according to speaker Tamer Sobeh, MD, of Tel Aviv University.
“The increasing workload for radiologists to detect [intracranial aneurysms]particularly incidental cases in the non-subarachnoid hemorrhage setting, emphasize the need for high-performance computer-assisted diagnostic tools to improve efficiency and increase sensitivity,” they wrote.
The researchers retrospectively applied a commercial deep learning system to 2,617 CTA cases acquired at their institution. Of these, 127 (4.7%) included an intracranial aneurysm.
The algorithm flagged 34 cases as suspected missed aneurysms, and 23 (67%) were considered by at least two out of three neuroradiologists to be true-positive aneurysms. This corresponds to a 23% increased detection rate, according to the researchers.
“The study demonstrates the potential of deep learning systems as a secondary reader to identify missed intracranial aneurysms, some of which could be clinically significant,” the authors concluded.
What else did they find? Stop by this Thursday session for more information.