This study focuses on developing an explainable machine learning model based on cardiopulmonary interaction characteristics to achieve early prediction of acute lung injury (ALI) in patients undergoing major liver surgery. The research will establish a digital early-warning system for ALI to provide support for clinical diagnosis and treatment decisions, thereby reducing the incidence and fatality rate of ALI.

This study will leverage cardiopulmonary interaction parameters to predict ALI in patients undergoing major liver surgery. Specifically, the research will collect data from preoperative, intraoperative, and postoperative phases. Machine learning algorithms-including logistic regression, random forest, support vector machines (SVM), and neural networks-will be used to develop and validate the prediction model. Model performance will be evaluated using metrics such as accuracy, sensitivity, specificity, and the receiver operating characteristic (ROC) curve. The ultimate objective is to develop a highly accurate and interpretable model that can be integrated into a digital early-warning system for clinical application.

Inclusion Criteria:

• Age ≥ 18 years
• Undergoing major liver surgery (including two-segment or more hepatectomy, liver transplantation, etc.)
• Voluntary participation with signed informed consent

• Patient refusal
• Presence of comorbidities affecting outcome assessment
• Incomplete outcome records
• Postoperative follow-up failure
• Participation in other studies that may interfere with outcome assessment