Enhancing post-induction hypotension prediction based on exemplar learning with crossover restart strategy driven...

Area of Science:

• Anesthesiology and Critical Care Medicine
• Artificial Intelligence in Healthcare
• Cardiovascular Medicine

Background:

• Post-induction hypotension (PIH) is a common complication in patients undergoing general anesthesia, particularly those with cardiovascular conditions or fluid management issues.
• PIH can lead to critical organ hypoperfusion, increasing risks of prolonged recovery, complications, and mortality.
• Early prediction of PIH is essential for optimizing patient management and improving surgical outcomes.

Purpose of the Study:

• To develop and validate a machine learning model for the accurate prediction of post-induction hypotension (PIH) in elderly patients undergoing elective surgery.
• To identify key clinical features that predict the occurrence of PIH.

Main Methods:

• A machine learning model, bECRIME-SVM, was developed using data from 440 elderly patients undergoing elective surgery.
• The model utilized an exemplar learning strategy with a crossover restart strategy within the rime optimization algorithm (ECRIME) for feature selection, followed by support vector machine (SVM) evaluation.
• Patients were classified into PIH and non-PIH groups based on mean arterial pressure post-induction.

Main Results:

• The bECRIME-SVM model achieved a prediction accuracy of 84.1% and a specificity of 85.3% for PIH.
• The ECRIME algorithm demonstrated superior optimization and convergence accuracy compared to other benchmark models.
• Key predictive features identified include diabetes, alcohol consumption, atropine use, beta-blocker use, total cholesterol, and pre-induction systolic blood pressure.

Conclusions:

• The bECRIME-SVM model is a valuable tool for accurate clinical prediction of PIH.
• Identifying significant predictive factors provides crucial insights for early detection and management of PIH.
• This predictive capability can lead to improved postoperative outcomes for patients receiving general anesthesia.