Unbiased Feature Selection in High Content Analysis of Biological Image Samples

摘要

A high content analysis computing system, computer program product and method provides receiving two or more sets of microscopic images of respective sites of selected biological cell assay. Each set of microscopic images comprises one or more images having an aggregate field of view sufficient to encompass all of the cells within the respective site and a resolution sufficient to image cellular bodies within each of encompassed cells. For each set of microscopic images, a library of algorithms is executed that evaluate image features for each cell in each site producing a quantified feature measurement for each cell. Quantified feature measurements are compared for each one of the two or more sets of microscopic images. A subset of the library of algorithms is identified that discriminate at least two of the two sets or more microscopic images and deployed as a computer program product for evaluating additional biological cell assays.