Related Experiment Video
Updated: Jul 12, 2025
12:18
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.6K
A risk prediction model based on machine learning for postoperative cognitive dysfunction in elderly patients with
Xianhai Xie1,2, Junlin Li1,2, Yi Zhong3
1School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, China.
Aging Clinical and Experimental Research
|October 21, 2023
Summary
This study developed a machine learning model to predict postoperative cognitive dysfunction (POCD) in elderly patients after non-cardiac surgery. The SVM model accurately identified high-risk patients, aiding in early intervention and prevention strategies.
Area of Science:
- Geriatric Medicine
- Surgical Oncology
- Artificial Intelligence in Healthcare
Background:
- Postoperative cognitive dysfunction (POCD) is a significant concern in elderly patients undergoing non-cardiac surgery.
- Early identification of at-risk individuals is crucial for implementing preventive measures and optimizing resource allocation.
Purpose of the Study:
- To develop a simple, clinically applicable machine learning (ML) model for predicting POCD at 3 months post-surgery in elderly patients.
- To enhance early detection and management strategies for POCD.
Main Methods:
- Utilized LASSO regression for feature selection and built five ML models to predict POCD risk.
- Employed Shapley Additive exPlanations (SHAP) for model interpretability.
- Collected data from 415 elderly patients undergoing non-cardiac surgery.
Main Results:
- The Support Vector Machine (SVM) model demonstrated superior performance among the five ML models.
- Key predictive features identified by SHAP include VAS score, age, intraoperative hypotension, and preoperative hemoglobin.
- The SVM model exhibited good interpretability and reliability in predicting POCD risk.
Conclusions:
- Successfully developed ML models, with SVM as the best performer, to predict POCD in elderly non-cardiac surgical patients.
- These models, based on six perioperative variables, can serve as decision aids for clinicians.
- The findings support closer monitoring and timely interventions for high-risk patients.
