Third, the need for multivariate analysis has long been known in the neuroimaging field—no serious neuroscientist would predict diagnosis from a single pixel or voxel in an image—the multivariate pattern of signals is paramount. Methods to access and recover the maximum amount of pertinent information in an image dataset are only just in their infancy in the fields of genetics and neuroimaging—partially due to the lack of sufficient data to test competing methods, until recently (Liu et al. 2009; Stein et al. 2010; Hibar et al. 2011b; Vounou et al. 2010; Le Floch et al. 2012; Rosenblatt et al. 2013; Meda et al. 2010, 2012).
