Personalized intracranial aneurysm rupture prognosis using Simulation-Based 4D Flow MRI and Machine Learning

Unruptured intracranial aneurysm (UIA) is a severe, relatively common cerebrovascular disorder in the general population.

Although such aneurysms are mostly asymptomatic and may not burst, a considerable number of UIA patients remain on a high risk of aneurysm rupture, a serious life-threatening condition. Thus, is crucial to decide the timing and type (clipping, coils, and/or stent, flow diverter) of surgical intervention.

However, the optimal management strategy of UIA is still open to clinical debate, with recommendations for elective repair after diagnosis based primarily on the aneurysm size and location.

While cardiovascular flow imaging has strong potential to provide the required information for surgical decision-making, currently, flow imaging and assessment is not sufficient to reliably predict rupture and quantitatively assess the risk of haemorrhage.

By identifying factors contributing to UIA rupture and by incorporating advanced computational techniques, i.e. 4D flow magnetic resonance imaging (MRI), machine learning (ML) algorithms, and computational fluid dynamics (CFD) models, this project will technologically advance quantification of rupture risk assessment on an UIA patient-specific basis.This project, Sim4DFlow, core technological aim is to develop a novel simulation-based imaging framework that can integrate the advanced computational techniques of 4D flow MRI, ML, and CFD.

Subsequently, Sim4DFlow will test and validate the framework prognostic capacity in the laboratory setting against relevant data of UIA patients.