Back to all papers

Chronic Subdural Hematoma Segmentation: A Dedicated Model to Overcome the Limitations of Acute Hemorrhage Segmentation Across Chronic Subdural Hematoma Subtypes and Density Variations.

April 14, 2026pubmed logopapers

Authors

Reddy B,Mutyam R,Fleiter TR,Dreizin D,Kamel PI

Affiliations (3)

  • Department of Diagnostic Radiology and Nuclear Medicine, University of Maryland School of Medicine, 22 S. Greene Street, Baltimore, MD, 21201, USA.
  • Interdisciplinary Program in Neuroscience, George Mason University, 4400 University Dr, Fairfax, VA, 22030, USA.
  • Department of Neuroradiology, Division of Diagnostic Imaging, The University of Texas MD Anderson Cancer Center, 1400 Pressler St., Unit 1479, Houston, TX, 77030, USA. [email protected].

Abstract

Most existing intracranial hematoma segmentation models target acute hemorrhages and may not generalize to the heterogeneous morphology of chronic subdural hematomas (CSDH). We compared a model trained on an open-access acute intracranial hemorrhage dataset with a model trained in combination with a CSDH dataset and further evaluated the performance of the combined dataset model across Nakaguchi subtypes of CSDH (homogeneous, laminar, separated, and trabecular) and between isodense and non-isodense hematomas. We analyzed 377 patients with 512 CSDHs. A 3D nnU-Net model was initialized with the open-access Brain Hemorrhage Segmentation Dataset (BHSD). In the second stage, the BHSD and institutional CSDH data (75% training, 25% testing) were combined for retraining. Finally, a model trained exclusively on the CSDH dataset was developed for comparison with the combined dataset model. Segmentation performance was evaluated using the Dice similarity coefficient (DSC) stratified by subtype and volume. The combined dataset model outperformed the BHSD-only model (mean DSC 0.917 ± 0.099 vs. 0.425 ± 0.306; P < .001) with greatest improvements in homogeneous (0.902 ± 0.072 vs. 0.350 ± 0.325; P < .001) and trabecular (0.945 ± 0.028 vs. 0.408 ± 0.266; P < .001) hematomas. No significant difference in DSC was observed between the combined dataset and the CSDH-only dataset models. The combined dataset model showed no significant difference in performance between isodense and non-isodense hematomas (P = 0.13). Segmentation performance correlated modestly with hematoma volume (Pearson r = 0.297; P < .001), with lower DSCs in smaller volumes (< 25 cm<sup>3</sup>: 0.786 ± 0.264 vs. > 25 cm<sup>3</sup>: 0.931 ± 0.046; P < .001). The proposed deep learning model enables more accurate and reliable segmentation of CSDH across hematoma subtypes and densities.

Topics

Journal Article

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.