Predictive Modeling of Brain Herniation Risk Factors and Critical Thresholds in Spontaneous intracerebral hemorrhage: A Pilot Study

Jinxiu Peng¹Shugang Wang²Jinpeng Wang³Bing Liu⁴Yimei Yuan⁴Lishan Yin^3*

• ¹Shandong Second Medical University, Weifang, Shandong Province, China
• ²Weifang Traditional Chinese Hospital, Weifang, Shandong Province, China
• ³Weifang People's Hospital, Weifang, China
• ⁴Affiliated Hospital of Weifang Medical University, Weifang, Shandong Province, China

Objective: Patients with spontaneous intracerebral hemorrhage(sICH) could benefit from personalized treatment strategies. Developing a rational classification system is therefore essential to guide clinical decision-making. This study aimed to identify independent predictors of brain herniation in sICH patients and establish critical thresholds for these predictors, using clinical and imaging data. Methods: We retrospectively analyzed consecutive spontaneous intracerebral hemorrhage patients admitted between June 2020 and December 2023. Demographics, medical history, clinical parameters on admission, and CT findings were collected. Hematoma volume and midline shift were quantified using 3D Slicer software, while intraventricular hemorrhage was graded by the Graeb score. Brain herniation was defined by acute neurological deterioration (e.g., loss of consciousness, anisocoria, or posturing) and CT evidence of critical structural displacement (e.g., obliteration of perimesencephalic cisterns, uncal herniation). Statistical methods included receiver operating characteristic curve analysis and multivariable binary logistic regression to identify independent predictors of herniation. Results: Fifty-five patients met inclusion criteria. Among them, 58 CT scans were analyzed. Multivariable analysis identified three independent predictors of cerebral herniation in basal ganglia/thalamic hemorrhages: hematoma volume >64 mL (adjusted OR = 14.67; 95% CI: 1.44–149.82; P =0.023), midline shift at the interventricular foramen >11 mm (adjusted OR = 10.05; 95% CI: 1.61–62.69; P = 0.014), and Graeb score (per 1-point increase: adjusted OR = 1.47; 95% CI: 1.08–2.00; P = 0.015). Conclusion: Among four midline structures analyzed, midline shift at the interventricular foramen was the strongest predictor. Key herniation predictors for basal ganglia/thalamic hemorrhage comprise hematoma volume >64 mL, midline shift at the interventricular foramen >11 mm, and Graeb score.