In a typical neuroimaging study—both now and 20 years ago—between ten and a few hundred subjects might have been scanned, and statistical models would be fitted to identify factors that affect brain structure and function. Early studies—such as lesion studies—correlated radiological measures with clinical diagnosis and behavior, but the study of large populations represented a new movement in human brain mapping. Fundamental questions in neuroscience could now be examined—what are the effects of aging, degenerative disease and psychiatric illness on the living brain? How do brain measures relate to cognition and behavior? Do brain measures predict our risk for disease, or prognosis in those who are ill?
