Comparison between gradients and parcellations for functional connectivity prediction of behavior

• Published in: NeuroImage (Orlando, Fla.) Fla.), 2023-06, Vol.273, p.120044-120044, Article 120044
• Main Authors: Kong, Ru, Tan, Yan Rui, Wulan, Naren, Ooi, Leon Qi Rong, Farahibozorg, Seyedeh-Rezvan, Harrison, Samuel, Bijsterbosch, Janine D., Bernhardt, Boris C., Eickhoff, Simon, Thomas Yeo, B.T.
• Format: Article
• Language: English
• Subjects: Adolescent; Algorithms; Behavior; Brain - diagnostic imaging; Cognitive ability; Cognitive development; Connectome - methods; Datasets; Humans; Image Processing, Computer-Assisted - methods; Magnetic Resonance Imaging - methods; Neural networks; Topography
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2023.120044

•Different parcellation and gradient approaches were compared for RSFC prediction of a broad range of behavioral measures in two datasets.•Across two regression algorithms (LRR and KRR), individual-specific hard-parcellation performed the best in the HCP dataset.•Principal gradients and all parcellation approaches performed similarly in the ABCD dataset.•Principal gradient approach required at least 40 to 60 gradients in order to perform as well as parcellation approaches. Resting-state functional connectivity (RSFC) is widely used to predict behavioral measures. To predict behavioral measures, representing RSFC with parcellations and gradients are the two most popular approaches. Here, we compare parcellation and gradient approaches for RSFC-based prediction of a broad range of behavioral measures in the Human Connectome Project (HCP) and Adolescent Brain Cognitive Development (ABCD) datasets. Among the parcellation approaches, we consider group-average “hard” parcellations (Schaefer et al., 2018), individual-specific “hard” parcellations (Kong et al., 2021a), and an individual-specific “soft” parcellation (spatial independent component analysis with dual regression; Beckmann et al., 2009). For gradient approaches, we consider the well-known principal gradients (Margulies et al., 2016) and the local gradient approach that detects local RSFC changes (Laumann et al., 2015). Across two regression algorithms, individual-specific hard-parcellation performs the best in the HCP dataset, while the principal gradients, spatial independent component analysis and group-average “hard” parcellations exhibit similar performance. On the other hand, principal gradients and all parcellation approaches perform similarly in the ABCD dataset. Across both datasets, local gradients perform the worst. Finally, we find that the principal gradient approach requires at least 40 to 60 gradients to perform as well as parcellation approaches. While most principal gradient studies utilize a single gradient, our results suggest that incorporating higher order gradients can provide significant behaviorally relevant information. Future work will consider the inclusion of additional parcellation and gradient approaches for comparison.