Funded Projects

Intracranial aneurysms (IAs) are potentially life-threatening vascular malformations, and the mainstay of modern treatment uses minimally invasive endovascular methods to occlude the lesion without the need for open surgery. Despite improvements in devices and approaches, up to 30% remain incompletely occluded, and thus at risk for rupture, at twelve months. The neurosurgery community currently lacks the ability to accurately identify which patients are at risk for poor outcomes at the time of presentation and treatment. Furthermore, after aneurysmal rupture, the largest source of morbidity and mortality is delayed cerebral ischemia due to vasospasm of the cerebrovasculature. Good clinical outcomes therefore require effective occlusion to prevent re-rupture, and early identification and aggressive management of vasospasm to prevent stroke. We plan to investigate the relationship between angiographic imaging biomarkers associated with contrast flow in Intracranial Aneurysms (IA) and treatment outcomes in order to develop a tool that improves our ability to predict success. The approach used to study these relations is to combine machine learning (ML) algorithms with Angiographic Parametric Imaging (API), also referred as colored Digital Subtraction Angiography. API imaging biomarkers and parametric maps will serve as input data and the treatment outcome will be used for output classification. Next we will study the correlation between API and vasospasm for ruptured IAs. Because API allows for both locally specific as well as global views of flow conditions, it is anticipated that quantitative assessment combined with trained ML algorithms will more accurately predict onset of vasospasm than qualitative human assessment. Early recognition and appropriate management of IAs addresses the objective of the UB to accelerate the field of stroke research and will help clinicians who manage these patients, including stroke neurologists, and neurosurgeons.