The ABCs of Accurate Volumetric Measurement of Cerebral Hematoma

Abstract

BACKGROUND AND PURPOSE: Both initial hematoma volume and hematoma growth are independent predictors of clinical outcomes and mortality among intracerebral hemorrhage patients. The purpose of this study was to evaluate the accuracy of different computed tomography image acquisition protocols and hematoma volume measurement techniques. METHODS: We used plastic and cadaveric phantoms to determine the accuracy of different volumetric measurement techniques. We performed both axial and spiral computed tomography scans with 0.75-, 1.5-, 3.0-, and 4.5-mm-thick transverse sections (with no gap). Different measurement techniques (planimetry, ABC/2, and 3D rendering) and different window width/level settings (I, 150/50 versus II, 587/-321) were used to assess generated errors in volumetric calculations. RESULTS: Both axial and spiral computed tomography scans yielded similar percent errors for different slice thicknesses and different measurement techniques. Comparison of different measurement techniques revealed a significant difference in measurement error only from the ABC/2 method as compared with 3D-rendering measurements (P<0.0001). The overall measurement error according to the ABC/2 method was further increased by approximately 8% for irregularly shaped hematomas (P=0.0004). A significant percent difference in measurement error was observed between window width/levels I and II for both planimetry (mean difference across all thicknesses, 1.91 ± 3.78, P=0.004) and Analyze software (mean difference across all thicknesses, 6.92 ± 7.29, P<0.0001) methods. CONCLUSIONS: A better understanding of the limitations that may affect measurement of hematoma volume is crucial in the assessment of hematoma volume, which is considered an independent marker of clinical outcome.