COMPUTED DETECTION AND QUANTITATIVE MORPHOMETRY OF ALZHEIMER SENILE PLAQUES

Senile plaques (SP) are the most characteristic neuropathologic lesion s of Alzheimer's disease (AD) and studies of plaque cortical distribut ion, density. and morphology may lead to new information about the ori gin and pathogenesis of this disease. We have developed an automated c omputer image analysis program to detect SP (including diffuse and mat ure forms) and to measure SP size, shape, and fractional area or load in digital micrographs of silver-stained tissue sections. The plaques are detected with adaptive thresholding. requiring no user interaction . Measures of SP size, morphology, and load are readily calculated fro m the pixel values in the detected SP features. These measurements are achieved accurately and exhaustively, and this method offers an alter native to manual SP counting. We demonstrate its application to 4 case s spanning the full range of the severity of the disease.