Prediction of Cognitive Progression in Individuals with Mild Cognitive Impairment Using Radiomics as an Improvement of the ATN System: A Five-Year Follow-Up Study

• DOI: 10.3348/kjr.2021.0323
• Publication Date: 2022-01-03T03:12:21Z
• Authors: Rao Song; Xiaojia Wu; Huan Liu; Dajing Guo; Lin Tang; Wei Zhang; Junbang Feng; Chuanming Li
• Keywords: Aged, 80 and over; Male; Amyloid beta-Peptides; tau Proteins; Peptide Fragments; 3. Good health; Neuroimaging and Head & Neck; 03 medical and health sciences; Cognition; 0302 clinical medicine; Alzheimer Disease; Disease Progression; Humans; Cognitive Dysfunction; Female; Biomarkers; Aged; Follow-Up Studies

To improve the N biomarker in amyloid/tau/neurodegeneration system by radiomics and study its value for predicting cognitive progression individuals with mild impairment (MCI). A group of 147 healthy controls (HCs) (72 male; mean age ± standard deviation, 73.7 6.3 years), 197 patients MCI (114 72.2 7.1 128 Alzheimer's disease (AD) (74 8.4 years) were included. Optimal A, T, biomarkers discriminating HC AD selected using receiver operating characteristic (ROC) curve analysis. model containing comprehensive information whole cerebral cortex deep nuclei was established to create a new biomarker. Cerebrospinal fluid (CSF) evaluated determine optimal or T biomarkers. All followed up until conversion at least 60 months. The predictive radiomics-based analyzed Kaplan-Meier estimates log-rank test. showed an ROC area 0.998 between HC. CSF Aβ42 p-tau proteins identified as biomarkers, respectively. For on continuum, isolated A+ indicator stability, while abnormalities N, separately simultaneously, indicated high risk progression. suspected non-Alzheimer's pathophysiology, T+ appearance N+ AD. We proposed improved that could help identify who are higher In addition, we clarified single A/T/N MCI.