Comparison of the ABC/2 formula with computer-assisted volumetry of ischemic cerebellar stroke.

Abstract

Current guidelines suggest surgical decompression for ischemic cerebellar stroke in case of significant mass effect. Recent research has aimed to identify a possible threshold for mass effect. However, a computer-assisted volumetry in acute setting is time consuming and impracticable, wherefore the aim of this study was to assess the accuracy and clinical applicability of the ABC/2 method in case of ischemic cerebellar stroke. Imaging data of 125 patients, including preoperative CT or MRI scans were used for volumetric analysis. The ABC/2 formula using scans in axial and coronal planes. BrainLab® Elements software was used for computer assisted volumetry by defining the region of interest allowing automated volumetric calculation. Measurements were conducted independently by blinded clinicians. Pearson correlation and Bland-Altmann test were used for statistical analysis. Among the 125 cerebellar infarctions analyzed, there was no statistical difference of mean infarct volume measurement between the ABC/2 formula and computer-assisted volumetry (16.6mL vs. 15.91mL; range 0.8-67.7mL; p = 0.76). The Spearman correlation test indicated a strong correlation between the two methods (r = 0.985, 95% CI: 0.979-0.990, p < 0.0001). Discrepancies were most notable in smaller infarction volumes (<10 mL), prompting a subgroup analysis. For infarct volume less than 10mL, the ratio of volumetric differences ranged from 47% to 60%, with absolute volume differences from -3-3 mL whereas the ratio ranged from -20% to 29%, with absolute volume differences from -6-8 mL in cases with infarct volume greater equal 10mL. ABC/2 formula shows a good correlation with computer-assisted volumetry. Consequently, it could serve as a fast and practical tool for estimating cerebellar infarct volume and aiding decision-making in clinical practice. However, the limitations and variability of the ABC/2 method, particularly for smaller infarcts, must be considered.