论文部分内容阅读
Objective To predict the incidence of anxiety in Alzheimers disease (AD) patients by using machine-learning models. Methods A large randomized controlled clinical trial was analyzed in this study, which involved AD patients and caregivers from 6 different sites in the United States. The incidence of anxiety in AD patients was predicted by backpropagation artificial neural networks and several machine learning models, including Bayesian Networks, logistic regression, ADTree, J48, and Decision table. Results Among all models for predicting the incidence of anxiety in AD patients, the artificial neural network with respectively 6 and 3 neurons in the first and second hidden layers achieved the highest predictive accuracy of 85.56 %. The decision tree revealed three main risk factors: "caregiver experiencing psychological distress", "caregiver suffering from chronic disease or cancer", and "lack of professional care service". Conclusion The unique ability of artificial neural networks on classifying nonlinearly separable problems may substantially benefit the prediction, prevention and early intervention of anxiety in Alzheimers patients. Decision tree has the double efficacy of predicting the incidence and discovering the risk factors of anxiety in Alzheimers patients. More resources should be provided to caregivers to improve their mental and physical health, and more professional care services should be adopted by Alzheimers families.