Abstract
Background and Purpose-Hematoma volume measurements influence prognosis and treatment decisions in patientswith spontaneous intracerebral hemorrhage (ICH). The aims of this study are to derive and validate a fully automatedsegmentation algorithm for ICH volumetric analysis using deep learning methods.Methods-In-patient computed tomography scans of 300 consecutive adults (age =18 years) with spontaneous, supratentorialICH who were enrolled in the ICHOP (Intracerebral Hemorrhage Outcomes Project; 2009-2018) were separated intotraining (n=260) and test (n=40) datasets. A fully automated segmentation algorithm was derived using convolutionalneural networks, and it was trained on manual segmentations from the training dataset. The algorithm's performance wasassessed against manual and semiautomated segmentation methods in the test dataset.Results-The mean volumetric Dice similarity coefficients for the fully automated segmentation algorithm when testedagainst manual and semiautomated segmentation methods were 0.894±0.264 and 0.905±0.254, respectively. ICH volumesderived from fully automated versus manual (R2=0.981; P<0.0001), fully automated versus semiautomated (R2=0.978;P<0.0001), and semiautomated versus manual (R2=0.990; P<0001) segmentation methods had strong between-groupcorrelations. The fully automated segmentation algorithm (mean 12.0±2.7 s/scan) was significantly faster than both ofthe manual (mean 201.5±92.2 s/scan; P<0.001) and semiautomated (mean 288.58±160.3 s/scan; P<0.001) segmentationmethods.Conclusions-The fully automated segmentation algorithm quantified hematoma volumes from computed tomographyscans of supratentorial ICH patients with similar accuracy and substantially greater efficiency compared with manualand semiautomated segmentation methods. External validation of the fully automated segmentation algorithm iswarranted. © 2019 American Heart Association, Inc.