Cerebral localization of emotion based on clinical-neuropathological correlations: methodological issues.

The method of clinico-pathological correlation for drawing inferences about the localization of particular cerebral functions has a long history of use, and well established theoretical limitations. Release phenomena, loss of excitatory drive, as well as non-specific tissue responses to injury may all have a bearing on observed behavioral change. Nevertheless, the consistent observation that severity of depression in stroke patients is greater for left hemispheric strokes, and greater for left frontal versus left occipital strokes is of considerable interest. Site of lesion appears to have greater explanatory power for this emotional symptom than the obvious psychological explanations in terms of loss of self-esteem and loss of function. Depression is greater for strokes in general than would be expected for equivalent loss of motor function with orthopedic etiology. Loss of cognitive function likewise is a poorer guide to severity of depression than site of lesion. On the other hand, accuracy of lesion assessment using present static anatomical methods (CAT scan), and reliability and validity of the psychopathological examination present methodological difficulties which are discussed. As newer brain imaging techniques that are sensitive to function are developed, this line of enquiry holds considerable promise for furthering our understanding of the anatomy and physiology of emotion.