An Automatic Cascaded Model for Hemorrhagic Stroke Segmentation and Hemorrhagic Volume Estimation
Weijin Xu, Zhuang Sha, Huihua Yang, Rongcai Jiang, Zhanying Li, Wentao Liu, Ruisheng Su
TL;DR
This study tackles rapid, accurate hemorrhagic stroke delineation and hemorrhage volume estimation on CT images. It introduces a cascaded 3D UNet with two-stage coarse-to-fine segmentation and deep supervision, achieving a $\text{DSC}$ of $85.66\%$ and $\text{IOU}$ of $76.82\%$ while reducing per-sample processing time to $6.2$ s. The method markedly improves volume estimation compared with the clinically used Tada formula $\frac{1}{2\times A\times B\times C}$, delivering substantial MAE improvements and faster results, which supports faster, quantitative clinical decision-making. Overall, the approach demonstrates strong segmentation performance and efficient volume estimation, enabling more quantitative CT assessments in stroke care.
Abstract
Hemorrhagic Stroke (HS) has a rapid onset and is a serious condition that poses a great health threat. Promptly and accurately delineating the bleeding region and estimating the volume of bleeding in Computer Tomography (CT) images can assist clinicians in treatment planning, leading to improved treatment outcomes for patients. In this paper, a cascaded 3D model is constructed based on UNet to perform a two-stage segmentation of the hemorrhage area in CT images from rough to fine, and the hemorrhage volume is automatically calculated from the segmented area. On a dataset with 341 cases of hemorrhagic stroke CT scans, the proposed model provides high-quality segmentation outcome with higher accuracy (DSC 85.66%) and better computation efficiency (6.2 second per sample) when compared to the traditional Tada formula with respect to hemorrhage volume estimation.