What question did this study set out to answer?

The aim is to develop and validate a machine-learning model for predicting postoperative complications after thoracoscopic mediastinal tumor surgery.

March 22, 2026Open Access

Interpretable machine learning model predicts postoperative complications after thoracoscopic mediastinal tumor surgery: a multicenter study

Key Points

The aim is to develop and validate a machine-learning model for predicting postoperative complications after thoracoscopic mediastinal tumor surgery.
Retrospective enrollment of 302 patients undergoing thoracoscopic resection from multiple centers.
Random division into training (70%) and validation (30%) sets.
Construction of fifteen predictive models using feature selection and machine learning algorithms.
Assessment of model performance using AUC and analysis of interpretability via SHAP.
Postoperative complications were observed in 43.6% of patients in the training set and 44.0% in the test set.
The Lasso-random forest model achieved the highest AUC of 0.799, indicating strong predictive performance.
Key features influencing complications included maximum tumor diameter and history of myasthenia gravis.

Abstract

Postoperative complications of mediastinal tumors significantly impact patients’ quality of life and long-term outcomes; however, a notable gap in the development of predictive tools for their occurrence remains. This study aimed to develop and validate a machine-learning model predicting thoracoscopic resection complications. Patients who underwent thoracoscopic mediastinal tumor resection at Southwest Hospital (January 2014 to April 2024) were retrospectively enrolled (n = 302) and randomly divided into training (70%) and validation (30%) sets. An additional 21 patients from Banan Hospital who underwent the same procedure (October 2023 to April 2024) were included as an external test set. The primary endpoint was postoperative complications within 90 days, with severe complications (Clavien-Dindo grade ≥Ⅱ) as the secondary endpoint. Fifteen predictive models were constructed using three feature selection methods and five machine learning algorithms. Model performance was assessed by AUC, and interpretability was analyzed using SHAP. The optimal model was selected based on the highest AUC values in validation set. Among the 302 patients in the main center, postoperative complications were observed in 92 (43.6%) in the training set and 40 (44.0%) in the test set. The Lasso-random forest model performed best, incorporating features like maximum tumor diameter, past medical history, surgical approach, myasthenia gravis, and hypertension (ranked by SHAP-derived feature importance). It achieved an AUC of 0.799 (95% CI: 0.700–0.897), showing robust discrimination and classification ability. The first web-based machine learning predictive model was developed to guide perioperative management and intraoperative decision-making for high-risk patients.

Bookmark

View Full Paper