New approaches to quantitative neuropathology: multivariate analysis of morphologic and neurochemical measures

The excellent review by Coleman and Flood on neuropathological changes in normal aging and Alzheimer's disease highlights the need for development and application of computer-assisted image analysis to the study of neurons in these conditions. The morphological and neurochemical changes in normal and pathological aging require quantitation and statistical analysis that can be best performed with the assistance of the image and data processing capabilities of the computer. Advanced image processing systems are being developed to identify and classify neurons according to several intelligently chosen visual features, apply discriminant analysis and population statistics to this data, and correlate this information to other neurochemical measurements as well as the clinical history of the patient. Techniques such as immunocytochemistry, receptor autoradiography and in situ hybridization produce information-rich images of the distribution of proteins and nucleic acids in tissue slices that can be analyzed by this approach.