Novel application of image analysis to the detection of spongiform change.

Abstract

Spongiform change is the characteristic neuropathologic abnormality within the brain in Creutzfeldt-Jakob disease, the most common human spongiform encephalopathy, but traditional methods of evaluating this abnormality are subjective and laborious. In order to overcome these difficulties, a quantitative method of assessment of spongiform change was devised and implemented using image analysis techniques. Successful validation of this system was performed by subjective and objective comparison with a neuropathologist's assessment. In a test-set of 50 microscopic images a strong positive rank correlation (Rs = 0.91) was established between the image analyzer and the neuropathologist. This system should facilitate large-scale analysis of different patterns of spongiform change in cases of human spongiform encephalopathy.