Study of AI Software for Detecting and Predicting Stability of Unruptured Cerebral Aneurysms Using Non-Invasive Imaging

What this Trial Is About

Aneurysmal subarachnoid hemorrhage SAH is a serious condition with a high mortality rate, and early diagnosis before rupture is critical to improving outcomes. With advances in non-invasive cerebrovascular imaging like CTA and MRA, more unruptured intracranial aneurysms UIAs are being detected. This study evaluates an artificial intelligence AI algorithm designed to automatically identify intracranial aneurysms from imaging data, aiming to improve diagnostic accuracy and help identify aneurysms at high risk of rupture to avoid unnecessary treatments and related harms. The study is a prospective, multicenter, registry observational study. It involves analyzing imaging data from patients undergoing different combinations of vascular imaging tests such as single or multiple MRAs, MRA combined with DSA, or MRA combined with CTA within three months. The AI software will be assessed by comparing its aneurysm detection and segmentation with manually annotated images and radiologists interpretations. Participants will undergo standardized non-invasive imaging follow-up over time. Researchers will collect imaging profiles and correlate them with aneurysm outcomes like growth or rupture. They will assess consistency between the AI model and expert annotations within one month. This approach aims to improve diagnosis and risk prediction for unruptured cerebral aneurysms, helping guide patient care based on imaging features and long-term follow-up data.

AI Models for Cerebral Aneurysms Segmentation, Detection and Stability Prediction