Table 1.

Minimum number of subjects (MNS) needed for the average graph measure to be within ± 5% of the value for the full group network. The numbers in bold text denote that discrimination between controls (CTR) and AD was achieved with less than 293 subjects at the given density, atlas and input measure.

Graph measure, density	Desikan thickness (CTR/AD)	Destrieux thickness (CTR/AD)	Desikan volume (CTR/AD)	Destrieux volume (CTR/AD)
Average strength, weighted	50/50	50/50	100/50	140/80
Global efficiency, 15%	135/50	150/50	130/50	50/50
Global efficiency, 25%	140/50	90/50	50/50	50/50
Global efficiency, weighted	50/50	90/50	130/90	170/140
Transitivity, 15%	120/50	125/60	120/75	180/95
Transitivity, 25%	105/50	135/55	140/50	160/85
Transitivity, weighted	50/50	50/50	60/50	130/50
Clustering, 15%	50/50	60/50	200/75	200/160
Clustering, 25%	50/50	100/50	135/60	180/150
Clustering, weighted	50/50	50/50	60/50	130/50
Char. path length, weighted	70/50	100/50	130/90	170/140