Uncovering Multi-Site Identifiability Based on Resting-State Functional Connectomes

TL;DR

This study presents a framework that enhances the reproducibility and individual identifiability of resting-state functional connectomes across multiple sites, addressing site variability issues in multi-site neuroimaging studies.

Contribution

The paper introduces a PCA-based method to improve multi-site identifiability of functional connectomes, significantly increasing reproducibility and biomarker potential.

Findings

Improved individual fingerprinting within and across sites.

Significant increase in ICC values for functional edges.

Optimal reconstruction maximizes differential identifiability regardless of fMRI volume count.

Abstract

Multi-site studies are becoming important to increase statistical power, enhance generalizability, and to improve the likelihood of pooling relevant subgroups together activities. Even with harmonized imaging sequences, site-dependent variability can mask the advantages of these multi-site studies. The aim of this study was to assess multi-site reproducibility in resting-state functional connectivity fingerprints, and to improve identifiability of functional connectomes. The individual fingerprinting of functional connectivity profiles is promising due to its potential as a robust neuroimaging biomarker. We evaluated, on two independent multi-site datasets, individual fingerprints in test-retest visit pairs within and across two sites and present a generalized framework based on principal component analysis to improve identifiability. Those components that maximized differential identifiability of a training dataset were used as an orthogonal connectivity basis to reconstruct the functional connectomes of training and validation sets. The optimally reconstructed functional connectomes showed a substantial improvement in individual fingerprinting within and across the two sites relative to the original data. A notable increase in ICC values for functional edges and resting-state networks was also observed. Improvements in identifiability were not found to be affected by global signal regression. Post-hoc analyses assessed the effect of the number of fMRI volumes on identifiability and showed that multi-site differential identifiability was for all cases maximized after optimal reconstruction. The generalizability of the optimal set of orthogonal basis of each dataset was evaluated through a leave-one-out procedure. Overall, results demonstrate that the framework presented in this study systematically improves identifiability in resting-state functional connectomes in multi-site studies.