Pattern recognition of multidimensional PBMC flow cytometry histograms for prostate cancer identification

Flow cytometry is a technique that is used to count cells and to characterize property of the cells. In spite of enormous information content on the cells provided by flow cytometry, cytometry data is still analyzed based on stepby-step gating, either manually or automatically via bioinformatics. This paper presents a new strategy of interpreting cytometry data in a different manner. The proposed strategy utilizes clustering approach to identify cell population of interest and supervised approach to identify statistical significant cell regions in the population that can differentiate prostate cancer patients from the benign patients.