为有效检测大脑是否出现快速老化以及早期预防、监测神经退化性疾病,本文基于磁共振弥散张量成像(DTI),经由大脑结构网络构建和图论分析,选取出与年龄有较强相关性的网络拓扑参数特征,进行多元线性回归建立了大脑老化预测模型。模型预测的脑年龄与受试者实际年龄的平均绝对误差为1.97岁,均方根误差为2.34岁。与同类模型相比,该模型具有良好的准确性和稳定性,基于此模型,可直接利用DTI影像来预测受试者的脑年龄。%In order to effectively detect the accelerated aging of brain and early prevent and monitor neurodegenerative diseases, this study selected network topology parameter features which were closely related to ages though the construction of the brain structural network and graph theory analysis based on diffusion tensor imaging (DTI). Then multiple linear regression of the features was used to establish a brain aging prediction model. The mean absolute error and the root mean square error of the model between the predicted age and actual age of the subject were 1.97 years and 2.34 years, respectively. Compared with other models, this model owns good accuracy and stability. Based on the application of the model, the brain age of the subject could be evaluated objectively according to their DTI images.
展开▼