Objective: This study aimed to explore the factors affecting the outcomes of spontaneous supratentorial cerebral hemorrhage 90 days after surgery. Methods: A total of 256 patients with spontaneous supratentorial intracerebral hemorrhage underwent craniotomy evacuation of hematoma. The control group included 120 patients who received conservative treatment. The patients were divided into two subgroups based on a bifurcation of the modified Rankin Scale (mRS) 90 days after clinical therapeutics: good outcome (mRS score 0–3) and poor outcome (mRS score 4–6). The differences in clinical and imaging data between the two subgroups were analyzed. Based on difference analysis results, a binary logistic regression model was constructed to analyze the influencing factors related to poor outcomes. Results: The difference analysis results in the surgery group showed statistically significant differences in age, sex, Glasgow Coma Score (GCS) on admission, coronary atherosclerosis, smoking, stroke history, blood glucose, D-dimer, hematoma size, deep cerebral hemorrhage, midline shift, hematoma burst into the ventricle, vortex sign, island sign, and black hole sign. Binary logistic regression analysis showed that deep cerebral hemorrhage, midline shift, and age > 58 years independently correlated with the poor outcomes of patients after surgery. The binary logistic regression results of the control group showed that age > 58 years and GCS ≤ 8 independently correlated with the poor outcomes of patients. Conclusions: Deep cerebral hemorrhage, midline shift, and age > 58 years significantly increased the risk of adverse prognosis in patients after surgery. The findings might help select the clinical treatment plan and evaluate the postoperative prognosis of patients.
This study aims to explore the value of a machine learning (ML) model based on radiomics features and clinical features in predicting the outcome of spontaneous supratentorial intracerebral hemorrhage (sICH) 90 days after surgery. A total of 348 patients with sICH underwent craniotomy evacuation of hematoma from three medical centers. One hundred and eight radiomics features were extracted from sICH lesions on baseline CT. Radiomics features were screened using 12 feature selection algorithms. Clinical features included age, gender, admission Glasgow Coma Scale (GCS), intraventricular hemorrhage (IVH), midline shift (MLS), and deep ICH. Nine ML models were constructed based on clinical feature, and clinical features + radiomics features, respectively. Grid search was performed on different combinations of feature selection and ML model for parameter tuning. The averaged receiver operating characteristics (ROC) area under curve (AUC) was calculated and the model with the largest AUC was selected. It was then tested using multicenter data. The combination of lasso regression feature selection and logistic regression model based on clinical features + radiomics features had the best performance (AUC: 0.87). The best model predicted an AUC of 0.85 (95%CI, 0.75–0.94) on the internal test set and 0.81 (95%CI, 0.64–0.99) and 0.83 (95%CI, 0.68–0.97) on the two external test sets, respectively. Twenty-two radiomics features were selected by lasso regression. The second-order feature gray level non-uniformity normalized was the most important radiomics feature. Age is the feature with the greatest contribution to prediction. The combination of clinical features and radiomics features using logistic regression models can improve the outcome prediction of patients with sICH 90 days after surgery.
Background: Understanding the risk factors for hematoma expansion (HE) in different regions of intracerebral hemorrhage (ICH) can help in the development of more accurate HE prediction tools and in implementing more effective clinical treatment interventions. This study aims to investigate the risk factors for HE in patients with lobar and deep ICH. Methods: A retrospective analysis was conducted on 558 cases of primary supratentorial ICH from Tongji Hospital Affiliated to Tongji University. Patients were categorized into lobar ICH and deep ICH groups. Differential analysis of ICH characteristics at different locations was performed, followed by subgroup analysis based on HE occurrence. Binary logistic regression was used to identify independent risk factors for HE in each group. Results: Among the 404 patients with ICH who underwent follow-up noncontrast computed tomography (NCCT) scans, the proportion with HE was similar in the deep ICH group (23.2%) and the lobar ICH group (22.7%). Binary logistic regression analysis revealed that fluid level (odds ratio [OR] 4.77, 95% confidence interval [CI] 1.74–13.06), admission Glasgow Coma Scale score (OR 0.87, 95% CI 0.80–0.96), and time from onset to NCCT examination (OR 0.84, 95% CI 0.75–0.94) were independently associated with HE in the deep ICH group. In the lobar ICH group, irregular shape (OR 4.96, 95% CI 1.37–18.01) and fibrinogen level (OR 0.42, 95% CI 0.21–0.86) were significant risk factors. Conclusions: Fluid level, low admission Glasgow Coma Scale score, and shorter time from onset to NCCT are independent predictors of HE in deep ICH, whereas irregular shape and low fibrinogen levels are independent predictors of HE in lobar ICH. These findings are of great significance for elucidating the mechanisms underlying HE in different locations of ICH and for developing precise predictive models of HE.
This study aimed to develop and validate a prognostic model integrating hematoma (R1), perilesional (R2), and clinical features to predict 90-day outcomes.