Study: AI boosts accuracy of CT angiography in spotting brain aneurysms

Nov. 3 (UPI) -- Artificial intelligence systems that use deep learning could help doctors detect cerebral aneurysms on CT angiography, according to a study published Tuesday by the journal Radiology.

The algorithm identified about 98% of cerebral aneurysms -- also called brain aneurysms -- and found eight new aneurysms that were overlooked on the initial assessment, the researchers said.

"The role of this deep-learning system, which has been trained to recognize aneurysms, is to give suggestions to the human reader to improve their performance and reduce mistakes," study co-author Xi Long said in a statement.

"The combined work of the human reader and computer system improves the diagnostic accuracy for the patient's sake," said Long, a professor of radiology at Tongji Medical College's Union Hospital in Wuhan, China.

A cerebral aneurysm is a weak or thin spot on an artery in the brain that balloons or bulges out and fills with blood, putting pressure on the nerves or brain tissue, according to the U.S. National Institute of Neurological Disorders and Stroke.

These aneurysms can burst or rupture, leaking blood into the surrounding tissue, causing a life-threatening hemorrhage, the institute said.

Detection of these aneurysms is critical, and CT angiography -- or computed tomography angiogram, a test that uses X-rays to provide detailed pictures of the brain and the blood vessels that go to it -- is usually the first choice for this process, Long and his colleagues said.

CT angiography is highly accurate, but cerebral aneurysms can be overlooked on the initial assessment due to their small size and the complexity of the blood vessels in the brain, they said.

For this study, the researchers developed a fully automated, highly sensitive algorithm for the detection of cerebral aneurysms on CT angiography images.

In their approach, the deep learning system is trained on existing images and learns to recognize abnormalities that can be difficult for a human observer to see.

The researchers used CT angiograms from more than 500 patients to train the deep learning system, and then they tested it on another 534 CT angiograms that included 649 aneurysms.

The algorithm detected 633 of the 649 cerebral aneurysms for a sensitivity of 97.5%, the data showed.

The results suggest that the deep learning algorithm has potential as a supportive tool for detecting cerebral aneurysms, as well as a potential to be used clinically for a second opinion during interpretation of CT angiography images, according to the researchers.

Still, it can miss very small aneurysms or aneurysms located close to similar density structures like bones. The method also also suffers from false positive results, meaning that it mistakenly identifies structures similar to aneurysms as aneurysms, they said.

"Simply put, the deep-learning system is intended to assist human readers, not to replace them," Long said.