 The study aimed to assess the feasibility of various machine learning models to pre-operatively predict the functional outcome of patients with chronic subdual hematoma, CSDH, and found that the support vector machine, SVM, model demonstrated the highest accuracy in external validation. This article was authored by Utaro Fuse, Yoshitoka Nagashima, Hiroshi Nishihaki, and others.