Correlations between hemodynamics and radiomic features in thrombosed intracranial aneurysms

TL;DR

This study finds that radiomic features of aneurysm wall enhancement correlate with hemodynamic factors like flow vortices and wall stress, offering new insights into thrombosed intracranial aneurysms.

Contribution

The study introduces a novel integration of radiomic analysis and hemodynamic modeling to explore intracranial aneurysm wall enhancement and thrombosis.

Findings

Six AWE radiomic features strongly correlate with aneurysm volume and surface area.

Ten AWE features are highly correlated with flow vortex parameters, and one with wall shear stress.

Saccular aneurysms show stronger AWE-WSS correlations, while fusiform aneurysms link AWE to vortex characteristics.

Abstract

Evaluating intracranial aneurysm (IA) rupture risk is essential for guiding management. Although intrasaccular thrombosis (IST) is less common, it can contribute to aneurysm growth, mass effect, and rupture. Aneurysm wall enhancement (AWE) on high-resolution MRI (HR-MRI) offers valuable insight into IST and IA progression. Using radiomics, we extracted spatial information of the aneurysm wall to characterize AWE. This study aimed to explore correlations between radiomics-based AWE profiles and gross hemodynamic parameters, integrating imaging and flow dynamics to better understand IST.

Radiomic analysis was conducted on a cohort of 3T HR-MRI scans from IA with IST. Three-dimensional vascular reconstructions and CFD simulations were conducted to quantify hemodynamic parameters. Spearman’s correlation was performed to correlate aneurysm morphology, AWE patterns, and aneurysmal hemodynamic characteristics.

A total of 37 thrombosed IAs were included in the analysis, comprising 22 fusiform (59.5%) and 15 saccular (40.5%) aneurysms. Six AWE RFs demonstrated strong correlations with aneurysm volume and surface area (ρ > 0.7 for both). Ten AWE RFs were highly correlated with flow vortex parameters (ρ > 0.7), and one showed a strong correlation with wall shear stress (WSS)-related metrics (ρ > 0.7). In the subset of saccular IAs, 20 AWE RFs were strongly associated with WSS-related metrics. In contrast, fusiform IAs showed stronger correlations between AWE RFs and vortex core characteristics. These findings suggest that elevated AWE is closely associated with regions of high oscillatory shear index and unstable flow vortices, indicating a potential link between wall enhancement and disturbed intra-aneurysmal hemodynamics.

Conclusions

Stagnant flow may promote degenerative remodeling of the aneurysm wall and IST. A combined spatiotemporal analysis of hemodynamic parameters and AWE patterns provide information about underlying biological processes of IAs, including the development of IST.

The online version contains supplementary material available at 10.1007/s00234-025-03730-x.

Figures