Traditional herbal medicine as an adjuvant therapy for prevent recurrence of hepatocellular carcinoma recurrence after radical resection: development and validation of a machine learning prediction model

Demographic characteristics included age and gender. Clinicopathological data encompassed the following elements: (1) etiology: hepatitis infection status and hepatitis infection duration; (2) tumor details: Edmondson-Steiner grade, pathological features, location, counts, size and Barcelona Clinic Liver Cancer (BCLC) stage; (3) liver function indicators: Child‒Pugh class and cirrhosis; (4) laboratory parameters: preresection and postresection (within two weeks) alpha-fetoprotein (AFP) levels, albumin (ALB), total bilirubin (STB), neutrophils (N), hemoglobin (HGB), and platelets (PLT); (5) treatment related factors: accepting THM intervention or not; (6) other data: platelet- albumin-bilirubin index (PALBI).