A DTI-Radiomics and Clinical Integration Model for Predicting MCI-to-AD Progression Using Corpus Callosum Features

This study aimed to explore the value of diffusion tensor imaging (DTI)- based radiomics in the early diagnosis of Alzheimer's disease (AD) and predicting the progression of mild cognitive impairment (MCI) to AD.

A cohort of 186 patients with MCI was obtained from the publicly accessible Alzheimer's. Disease Neuroimaging Initiative (ADNI) database, and 49 of these individuals developed AD over a 5-year observation period. The subjects were divided into a training set and a test set in a ratio of 7 to 3. Radiomic features were extracted from the corpus callosum within the DTI post-processed images. The Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression algorithm was employed to develop radiomic signatures. The performance of the radiomic signature was assessed using receiver operating characteristic (ROC) analysis and decision curve analysis (DCA).

In the training set, 35 patients were converted, and in the test set, 14 patients were converted. Among all the patients, notable differences were observed in age, CDR-SB, ADAS, MMSE, FAQ, and MOCA between the stable group and the transformed group (p < 0.05). In the test set, the AUCs of the radiomics signatures constructed based on fractional anisotropy, axial diffusivity, mean diffusivity, and radial diffusivity were 0.824, 0.852, 0.833, and 0.862, respectively. The AUC of the clinical model was 0.868, and that of the combined model was 0.936. DCA demonstrated that the combined model had the best performance.

The combined radiomics and clinical model, utilizing DTI data, can relatively accurately forecast which patients with MCI are likely to progress to AD. This approach offers potential for early AD prevention in MCI patients.