Related Experiment Video
Updated: Jul 14, 2026
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Development of a Postoperative Delirium Risk Prediction Model for Elderly Patients by Integrating MRI Radiomics With
Li Yang1, Jin Zhang1, Beiping Li1
1Xuzhou No. 1 People's Hospital, Jiangsu, China.
Purpose:
This study aims to explore the potential of magnetic resonance imaging (MRI) radiomics in assessing the risk of postoperative delirium in elderly patients and to construct an integrated postoperative delirium risk prediction model by integrating clinical characteristics.
Design:
A prospective cohort study.
Methods:
Three hundred elderly patients aged 65 and above who underwent elective general anesthesia surgery at the First People's Hospital of Xuzhou from January 2022 to December 2023 were prospectively included. MRI features were collected using a 3.0T Siemens magnetic resonance (MR) scanner, including hippocampal volume, intracranial volume, and diffuse white matter hyperintensity volume, along with clinical data. Statistical analysis was performed using R 4.2.2 software to construct and validate the predictive model.
Findings:
ASA level III or above, intraoperative hypotension, surgical duration, age, history of diabetes, postoperative VAS score, hippocampal volume, intracranial volume, and diffuse white matter hyperintensity volume were significantly correlated with postoperative delirium. Multivariate analysis revealed that age, surgical duration, postoperative VAS score, hippocampal volume, white matter hyperintensity volume, and perivascular space dilation were independent risk factors for postoperative delirium. The constructed postoperative delirium risk prediction model showed good predictive ability in both the training and validation sets, with ROC curve areas of 0.973 and 0.966, respectively, and high sensitivity and specificity.
Conclusions:
The postoperative delirium nomogram prediction model developed in this study shows good predictive and validation effects and can serve as a clinical decision support tool to help clinicians identify high-risk patients, implement preventive measures, and reduce the incidence of postoperative delirium.