Objective Quantification of Aneurysm Wall Enhancement Can Improve Experts’ Enhancement Adjudication.

Purpose: Aneurysm Wall Enhancement (AWE) has been used in some instances to determine which aneurysm rupture or to triage patients for treatment. The determination of AWE is mainly subjective. The accuracy of AWE adjudication can be improved by using AWE objective quantification methods.

Methods/Materials: Saccular aneurysms were imaged with high-resolution 3T MRIs (HR-MRI) between May 2019 and June 2023. Two neuroradiologists determined the presence or absence of AWE. HR-MRIs were processed to generate three-dimensional aneurysmal segmentations using 3D Slicer. Signal intensity (SI) distribution across the aneurysm wall was calculated utilizing an in-house MATLAB script. The average SI of the corpus callosum was utilized for normalization. The mean SI in the post-contrast T1 HR-MRI was used to calculate the 3D circumferential AWE (CAWE). An aneurysm was objectively defined to be “enhancing” when the CAWE value was higher than one. Morphological features such as size, aspect ratio, and size ratio were measured. Logistic regression was used to assess how aneurysm morphology and symptomatic presentation influenced the subjective adjudication of AWE.

Results: One hundred and thirteen aneurysms were included in the study. Radiologists subjectively classified 26 (23%) aneurysms as “enhancing” and 87 (77%) as “non-enhancing”. Conversely, our objective method identified 19 (17%) CAWE-positive and 94 (83%) CAWE-negative aneurysms. Subjective experts’ adjudication was different from the AWE objective adjudication (p= 0.0057). Among 26 aneurysms subjectively adjudicated as “enhancing”, only 9 were objectively CAWE positive (PPV: 34.61%). Out of 87 aneurysms classified as “non-enhancing”, 77 were CAWE negative (NPV: 88.51%). Expert’s enhancement adjudication was more likely to agree with the objective presence of AWE in symptomatic aneurysms and those with a higher aspect ratio (OR: 4.17, 95% CI 1.03 – 16.75, p= 0.0444, OR: 1.92, 95% CI 1.06 - 3.43, p=0.0295, respectively).

Conclusions: The presence of AWE tends to be overestimated when assessed subjectively. Objective quantification methods of AWE can improve the accuracy of AWE adjudication.