Table 2.

Group Differences in Global Network Metrics.

Network Metrics	p-value
	W i j =FA*FN/Volume
	ALL-90					HOA-1024	ALL-1024
	FN≥1	FN≥2	FN≥3	FN≥5	FN≥10	FN≥1	FN≥1
σ	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001
γ	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001
λ	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.
Cp	<0.001	<0.001	<0.001	<0.001	<0.001	0.001	0.003
Lp	n.s.	n.s.	n.s.	n.s.	n.s.	0.011	0.022
Eg	0.013	0.014	0.012	0.012	0.011	0.009	0.020
Eloc	0.016	0.015	0.017	0.018	0.017	n.s.	n.s.

The structural networks were construct by using the definitions of W _{i j} =FA*FN / Volume. The threshold of FN ≥ 1, or FN ≥ 2, or FN ≥ 3, or FN ≥ 5, or FN ≥ 10 indicates that two regions were connected if at least 1 or 2 or 3 or 5 or 10 streamlines exist between a pair of brain regions. The high spatial resolution template (HOA-1024 and ALL-1024) was also used in constructing the structural networks. The comparisons of network metrics between groups were performed by using a nonparametric permutation test while removing the effects of age, sex, and years of education. (P< 0.05; Benjamin-Hochberg FDR corrected) indicated a significant group difference.

FA, fractional anisotropy; FN, streamline number; σ, small-worldness; γ, normalized clustering coefficient; λ, normalized shortest path length; Cp, cluster coefficient; Lp, characteristic path length; Eg, global efficiency; Eloc, local efficiency; n.s., non-significant.