Identifying canonical and replicable multi‐scale intrinsic connectivity networks in 100k+ resting‐state fMRI datasets

• Published in: Human brain mapping Vol. 44; no. 17; pp. 5729 - 5748
• Main Authors: Iraji, A., Fu, Z., Faghiri, A., Duda, M., Chen, J., Rachakonda, S., DeRamus, T., Kochunov, P., Adhikari, B. M., Belger, A., Ford, J. M., Mathalon, D. H., Pearlson, G. D., Potkin, S. G., Preda, A., Turner, J. A., van Erp, T. G. M., Bustillo, J. R., Yang, K., Ishizuka, K., Faria, A., Sawa, A., Hutchison, K., Osuch, E. A., Theberge, J., Abbott, C., Mueller, B. A., Zhi, D., Zhuo, C., Liu, S., Xu, Y., Salman, M., Liu, J., Du, Y., Sui, J., Adali, T., Calhoun, V. D.
• Format: Journal Article
• Language: English
• Published: United States John Wiley & Sons, Inc 01.12.2023
• Subjects: Brain - diagnostic imaging; Brain mapping; Brain Mapping - methods; Datasets; Feasibility studies; Functional magnetic resonance imaging; Humans; Independent component analysis; Magnetic Resonance Imaging - methods; Medical imaging; Nerve Net - diagnostic imaging; Neural networks; Neuroimaging; Reproducibility of Results; Similarity; Smoothness; Spatial data; Spatial discrimination; Spatial resolution; functional connectivity (FC); functional templates; independent component analysis (ICA); intrinsic connectivity networks (ICNs)
• ISSN: 1065-9471; 1097-0193; 1097-0193
• DOI: 10.1002/hbm.26472

Despite the known benefits of data‐driven approaches, the lack of approaches for identifying functional neuroimaging patterns that capture both individual variations and inter‐subject correspondence limits the clinical utility of rsfMRI and its application to single‐subject analyses. Here, using rsfMRI data from over 100k individuals across private and public datasets, we identify replicable multi‐spatial‐scale canonical intrinsic connectivity network (ICN) templates via the use of multi‐model‐order independent component analysis (ICA). We also study the feasibility of estimating subject‐specific ICNs via spatially constrained ICA. The results show that the subject‐level ICN estimations vary as a function of the ICN itself, the data length, and the spatial resolution. In general, large‐scale ICNs require less data to achieve specific levels of (within‐ and between‐subject) spatial similarity with their templates. Importantly, increasing data length can reduce an ICN's subject‐level specificity, suggesting longer scans may not always be desirable. We also find a positive linear relationship between data length and spatial smoothness (possibly due to averaging over intrinsic dynamics), suggesting studies examining optimized data length should consider spatial smoothness. Finally, consistency in spatial similarity between ICNs estimated using the full data and subsets across different data lengths suggests lower within‐subject spatial similarity in shorter data is not wholly defined by lower reliability in ICN estimates, but may be an indication of meaningful brain dynamics which average out as data length increases.