Functional neuroimaging refers to the use of non-invasive brain imaging methods to localize neural activity in the human brain related to the engagement of specific mental functions. Such methods have been in use since the late 1980s to investigate the brain regions activated by sensory, motor and cognitive processing in healthy subjects and departures from normal patterns of activation in cases of neurological or psychiatric illness. Functional neuroimaging studies are an important complement to other approaches to understanding brain function, such as the neuropsychological (see NEUROPSYCHOLOGY) investigation of the effects of BRAIN INJURY in humans, and experimental studies of brain function in animals.
The two most important functional neuroimaging methods are POSITRON EMISSION TOMOGRAPHY (PET), which came into use in the mid-1980s, and FUNCTIONAL MAGNETIC RESONANCE IMAGING (fMRI), which was introduced approximately a decade later. Although differing in several important respects the two methods share one important feature, which is that with neither method is neural activity detected directly. Instead, the methods take advantage of the fact that neural activity and blood flow in the normal brain are closely coupled, such that within a few seconds, a change in activity triggers a concomitant change in flow. Thus, if an experimental manipulation has an effect on the net level of neural activity in a given brain region, the effect will be detected as a change in blood flow through that region. It is important to note that changes in the pattern of neural activity in a neural population, rather than in its overall level, will not be accompanied by changes in flow, and thus cannot be detected by PET or fMRI.
The principal goal of functional neuroimaging studies is to identify brain regions in which neural activity is associated selectively with a specific psychological function, such as COLOUR VISION OF CONSCIOUS RECOLLECTION, and thereby to identify the regions that support such functions. Typically, this goal is achieved by obtaining functional images during psychological tasks which have been designed to engage the psychological function or functions of interest. An image of the distribution of blood flow or oxygenation obtained during a single experimental task is, however, impossible to interpret, as there is no means of distinguishing between regions in which activity is task-specific, and regions in which it is non-specific.
Hence the general experimental strategy in functional neuroimaging studies is to compare images obtained in at least two experimental conditions. In the simplest experimental design, sometimes called the subtraction method, two conditions are employed, a baseline and a task condition. These conditions are designed to make equivalent demands upon sensory and cognitive processes, with the exception of the specific function of experimental interest, which is engaged in the task condition only. Because all other factors are held constant, any differences in the images obtained between the baseline and task conditions can be ascribed to the function of interest. The subtraction method is predicated on a number of key assumptions about the relationship between different cognitive functions which are not always met, and is limited in the range of experimental designs to which it is applicable. More complex experimental designs exist which overcome these problems (see Frith & Friston, 1997).
