Clear Detection of Thin-Walled Regions in Unruptured Cerebral Aneurysms by Using Computational Fluid Dynamics

OBJECTIVE: Thin-walled regions (TIWRs) within cere-bral aneurysms have a high risk of rupture during surgicalmanipulation. Previous reports have demonstrated specificchanges in the parameters of computational fluid dynamicsin TIWRs; however, they have not been fully evaluated. Weidentified and investigated a novel parameter, wall shearstress vector cycle variation (WSSVV), with user-friendlysoftware that could predict TIWRs. METHODS :Twelve unruptured cerebral aneurysms wereanalyzed. TIWRs were defined as reddish areas comparedwith the normal-colored parent artery on intraoperativeviews. The position and orientation of these clinical im-ages were adjusted to match the WSSVV color maps.TIWRs and thick-walled regions (TKWRs) were marked andcompared with the corresponding regions on WSSVV maps.The default images obtained from WSSVV imaging requiredappropriate maximum color bar value (MCBV) adjustmentfor predicting TIWRs. Sensitivity and specificity analyseswere performed by changing the MCBV from 300 to 700 atintervals of 100. With the optimal MCBV, the WSSVV valueswere quantitatively compared. RESULTS: All of the selected 18 TIWRs and 16 TKWRscorresponded to low- and high-value regions of the WSSVVcolor maps at the adjusted MCBV, respectively. The meanoptimal MCBV was 483.3 167.50 (range, 300e700). Accordingto receiver operating characteristic analysis, the best MCBVfor predicting TIWRs was 500 (highest sensitivity, 0.89; spec-ificity, 0.94). Under this condition, the quantitative values of thecomputational fluid dynamics color maps for TIWRs andTKWRs were significantly different (P