Test-retest reliability of computational network measurements derived from the structural connectome of the human brain

Abstract

Structural magnetic resonance (MR) connectomics holds promise for the diagnosis, outcome prediction, and treatment monitoring of many common neurodevelopmental, psychiatric, and neurodegenerative disorders for which there is currently no clinical utility for MR imaging (MRI). Before computational network metrics from the human connectome can be applied in a clinical setting, their precision and their normative intersubject variation must be understood to guide the study design and the interpretation of longitudinal data. In this work, the reproducibility of commonly used graph theoretic measures is investigated, as applied to the structural connectome of healthy adult volunteers. Two datasets are examined, one consisting of 10 subjects scanned twice at one MRI facility and one consisting of five subjects scanned once each at two different facilities using the same imaging platform. Global graph metrics are calculated for unweighed and weighed connectomes, and two levels of granularity of the connectome are evaluated: one based on the 82-node cortical and subcortical parcellation from FreeSurfer and one based on an atlas-free parcellation of the gray-white matter boundary consisting of 1000 cortical nodes. The consistency of the unweighed and weighed edges and the module assignments are also computed for the 82-node connectomes. Overall, the results demonstrate good-to-excellent test-retest reliability for the entire connectome-processing pipeline, including the graph analytics, in both the intrasite and intersite datasets. These findings indicate that measurements of computational network metrics derived from the structural connectome have sufficient precision to be tested as potential biomarkers for diagnosis, prognosis, and monitoring of interventions in neurological and psychiatric diseases.

FIG. 1.

Cartoon depicting the 82-node connectome-processing pipeline. First, the diffusion and structural magnetic resonance images are aligned; then, the FreeSurfer labels are transformed to the diffusion space to seed the tractography. Tractography is performed with each of the 82 seed regions to construct the structural connectome.

FIG. 2.

(a) Representative connectome, (b) representative module partition, and (c) mean degree distribution across 10 intrasite subjects.

FIG. 3.

Representative high-resolution cortical connectome: degree and strength of nodes.

FIG. 4.

Selected unweighed and weighed metrics: K, L, C, K^w, L^w, and C^w are plotted for the 82-node connectomes across the intrasite and intersite scans.

FIG. 5.

Selected unweighed and weighed metrics: K, L, C, K^w, L^w, and C^w are plotted for the high-resolution cortical connectomes across intrasite and intersite scans.

FIG. 6.

The effect of threshold on intraclass correlation coefficient (ICC) and percentage coefficient of variation (CV%) for the graph metrics computed for the 82- and 1000-node connectomes.

FIG. 7.

The effect of number of streamlines initiated from every seed voxel on ICC for the unweighed and weighed graph metrics computed for the 82- and 1000-node connectomes.

FIG. 8.

Edge consistency for binary and weighed connectomes: the edge agreement and correlation coefficient are plotted for pairs of intrasite and intersite scans. In each plot, the black points denote the intrasubject variability, and the red line denotes the mean intersubject variability with a dotted line showing the standard deviation (SD).

FIG. 9.

Consistency of individual connection weights: (a) Histogram of the CV% for all suprathreshold connection weights and (b) correlation of CV% of node-to-node connection weights and the number of tractography steps between the two nodes

FIG. 10.

Module assignment consistency: the mean Hubert Rand index (HRI) is plotted for (a) intrasite, intrascan data to measure the reliability of the community detection algorithm, (b) intrasite, interscan, and (c) intrasite, interscan. In (b) and (c), the black points denote the intrasubject variability with SD error bars, and the red line denotes the mean intersubject variability with a dotted line for the SD.