FIGURE 1.

Analytic pipeline. Step1: The time-course signal of 53 ICNs have been identified using group-ICA in the Neuromak template. Step2: After identifying 53 ICNs, a taper sliding window was used to segment the time-course signals and then calculated the functional network connectivity (FNC). Each subject has 139 FNCs with a size of 53 × 53. Also, we calculated static FNC for the entire time of recording. Step3: After vectorizing the FNC matrixes, we have concatenated them, and then a k-means clustering with correlation as distance metrics was used to group FNCs to three distinct clusters. Step4: Then, based on the state vector, we calculated between-state transition probability or hidden Markov model (HMM) features and occupancy rate (OCR) for each subject. In total, nine HMM features and three OCR were estimated from the state vector of each subject. Step5: To find a link between FNC features, including sFNC and dFNC feature with clinical dementia rating scale sum of boxes (CDR-SOB), we used partial correlation by accounting for age, gender.