Predictive model effectively identifies surgical margin risks in esophageal cancer.

A pathology-based predictive model for positive surgical margins (PSM) in esophageal squamous cell carcinoma (ESCC) has been established and validated. Analysis of 1,776 patients revealed that multifocal lesions, vascular invasion, and pathomics-based features significantly predict PSM. The decision tree model displayed excellent discrimination with an area under the ROC curve of 0.899, while the nomogram showed slightly lower performance at 0.803. This model may enhance preoperative evaluation and treatment strategies for ESCC patients.