多模态头颅MRI参数联合随机森林算法构建老年衰弱的早期预测模型OA

Objective:To construct an early prediction model for frailty in older adults using multimodal cranial MRI parameters combined with a random forest algorithm.Methods:A total of 500 elderly hospitalized patients who received treatment from January to December 2023 were selected as the research subjects.According to the frailty assessment scale,there were 140 patients in the frailty group,210 patients in the pre-frailty stage,and 150 patients in the non-frailty group.The differences between high white matter signals and deep white matter signals in patients were studied,and the differ-ences in lifestyle,multi-drug use,daily living ability,anxiety,and other factors among the three groups of patients were compared.Results:There were statistically significant differences(P＜0.05)in age,comorbidities,living conditions,marital sta-tus,education level,age,mini nutritional assessment short form,mini mental state examination,global deterioration scale 1,activities of daily living,types of chronic disease comorbidities,medication types,white matter hyperintensities,deep white matter hyperintensities,hippocampal volume,left hippocampal volume,and right hippocampal volume among patients in the frailty group,pre-frailty group,and non-frailty group.Multivariate analysis showed that the above factors were the contribut-ing factors to the patient's frailty.The model formed by the study was used in the construction of a random forest model.The validation set(30%data)evaluation has good predictive ability.Conclusion:The early prediction model for frailty in older adults,constructed using multimodal cranial MRI parameters combined with a random forest algorithm,effectively pre-dicts frailty and provides an important reference for early clinical intervention and frailty identification.