Expert system for flow cytometry data analysis: a preliminary report

Flow Cytometry has become an accepted technique in the clinical laboratory for rapid immunophenotyping of patient blood samples. Multiple, fluorescent labeled monoclonal antibodies are used to tag the cells, which are then analyzed one at a time at rates of several thousand cells a second. Patient samples are processed through the flow cytometer at more than one a minute. Clinicians are being overwhelmed by the large amount of data that must be analyzed to provide the information needed to assist in disease diagnosis. An expert system is being developed to assist clinicians in analyzing this multivariate flow cytometry data. The data from each sample are processed by a clustering algorithm, which finds the means of the distinct cell subpopulations in a sample. These mean values of fluorescence are translated into words such as "negative", "dim" and "bright" and the words are combined into patterns that are matched against the premises on the left hand side of the rules used to identify the disease categories. This is a report of work in progress.© (1990) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.