But a much more adventurous goal provided the driving force behind the new and emerging fields of imaging genomics. This goal was to use neuroimaging data directly, to screen the genome for common variants that might affect the brain. In other words, rather than using the images in secondary studies of what disease risk genes do, images could be screened to discover important genetic associations. (Instead of imaging genetics, the somewhat interchangeable term “imaging genomics” is also used; genomics tends to refer to any method that directly assesses variation in the genome, as opposed to studies that may assess a single locus only, or simpler family studies that may not even analyze DNA). The growing computational power to screen very large neuroimaging datasets—for the purpose of extracting meaningful features from them—made this an interesting and achievable objective. Advocates of “imaging genetics”—the genetic analysis of brain images (Glahn et al. 2007; Turner et al. 2006)—suggested that it might even be more efficient to screen traits derived from brain images to provide endophenotypes for brain disorders.
