A method for predicting mortality in acute mesenteric ischemia: Machine learning

Area of Science:

• Medical Informatics
• Computational Biology
• Clinical Medicine

Background:

• Acute mesenteric ischemia (AMI) presents a significant challenge in clinical practice.
• Predicting hospital mortality in AMI patients is crucial for timely intervention.
• Existing methods for mortality prediction in AMI may lack efficiency and accuracy.

Purpose of the Study:

• To develop and validate an artificial intelligence (AI) model using machine learning (ML) for predicting hospital mortality in patients with AMI.
• To identify independent risk factors associated with mortality in AMI.
• To establish an efficient and rapid method for mortality prediction in AMI.

Main Methods:

• A cohort of 122 patients with AMI was retrospectively analyzed.
• Patients were divided into training (n=97) and validation (n=25) cohorts.
• Serum laboratory results were used as features, with Recursive Feature Elimination (RFE) for feature selection.
• Machine learning algorithms including logistic regression, random forest, k-nearest neighbor, multilayer perceptron, support vector classifier, and a voting classifier were implemented in Python.

Main Results:

• The overall in-hospital mortality rate was 50%.
• Key predictors identified included age, red cell distribution width (RDW), C-reactive protein (CRP), lactate, globulin, and creatinine.
• The Support Vector Classifier (SVC) and Voting Classifier (VC) demonstrated the highest performance, achieving an 84% success rate, with SVC showing an Area Under the Curve (AUC) of 0.90.

Conclusions:

• Independent risk factors for mortality in AMI patients were successfully identified.
• An efficient and rapid ML-based method for predicting hospital mortality in AMI has been developed.
• The developed model shows promise for improving patient outcomes through early risk stratification.