The visualization of brain structure through computed tomography (CT) and magnetic resonance imaging (MRI) has greatly advanced our knowledge of psychiatric and neuropsychiatric diseases. Further and more refined information about structural and developmental abnormalities in psychiatric disorders can be expected from recent advances in structural MRI such as diffusion-weighted imaging (DWI) and magnetization transfer imaging (MTI), which have already been applied to the investigation of white matter defects in, for example, Alzheimer's disease (AD) [38,39] and schizophrenia [34]. The metabolic changes that underlie or accompany neuropsychiatric diseases, on the other hand, can be probed with the traditional radiotracer methods and with magnetic resonance spectroscopy (MRS), a technique that allows the non-invasive study of in vivo neurochemistry and has already gained a place in the differential diagnosis and the monitoring of the metabolic effects of drug treatment in a number of diseases [14,35,48,81,83]. However, if we want to find out how these changes bring about the respective neuropsychiatric diseases, we need to integrate the clinical pictures and structural and metabolic brain imaging with information about deficits in brain function that can, at present, only be obtained through the methods of nuclear medicine and non-invasive neurophysiology and neuroimaging [75]. The large bodies of studies of abnormal brain function with radiotracer methods (for review see, e.g., [24,81,88]) or non-invasive electrophysiology (for review see, e.g., [45]) in psychiatric patients fall outside the scope of this paper. They do, however, provide the background for the more recent attempts to visualize normal brain function and functional abnormalities using functional MRI (fMRI) based on the blood oxygenation-dependent (BOLD) effect [29], and most neuropsychiatric fMRI presuppose the findings of earlier radiotracer or electroencephalographic studies. While BOLD fMRI is a particularly promising tool for the study of mental illness, because it can be applied and repeated easily, is widely available, and lends itself to the study of individual functional anatomy [22], many groups working in the field of psychiatric neuroimaging apply more than one technique, and multimodal neuroimaging, combining fMRI with other functional, metabolic, and structural techniques, will be required for a true understanding of the way morphological and neurochemical abnormalities shape the clinical picture of neuropsychiatric disorders [59]. In this paper, we will present the state of fMRI research into the two most widespread neuropsychiatric disorders, schizophrenia and Alzheimer's disease. These disorders have, owing to their multi-faceted neurobiology, so far eluded any attempts at defining unequivocal aetiologies and pathomechanisms. At the same time, they present major public health and socioeconomic issues. It is therefore not surprising that they have attracted the attention of the neuroimaging community in a particular manner. (C) 2002 Elsevier Science B.V. All rights reserved.
