Expert Assistance for Flow Immunophenotyping
There is a serious data analysis bottleneck for flow immunophenotyping in the clinical laboratory when many parameters are required. Multiple tubes of cells are measured on the flow cytometer for each patient peripheral blood or bone marrow specimen. Each tube requires only a few minutes to measure five variates on each of 7000 or more cells. The data for each patient, however, may require 20 to 40 minutes to analyze with the manual methods now in common use. This paper describes the use of cluster analysis to replace the manual gating methods used to find the leukemic or other abnormal cell populations. It then discusses the use of an expert system to assist in describing the immunophenotype of the specimen. Since the literature on expert systems in cytometry has been reviewed recently (Salzman and Bartels, 1992), this article does not include such a review.
Unable to display preview. Download preview PDF.
- Davies DL, Boulton DW (1979) A cluster separation measure. IEEE Transactions on Pattern Analysis and machine Intelligence PAMI-1: 224–227.Google Scholar
- Everitt B (1986) Cluster analysis. 2nd edn. John Wiley & Sons, New York.Google Scholar
- Giarratano J, Riley G (1989) Expert Systems. Principles and Programming. P WS-Kent Pub. Co. Boston. CLIPS is available from the NASA software distribution center, COSMIC, 382 E. Broad Street, Athens, GA 30602 USA. (404)542–3265.Google Scholar
- Salzman GC, Bartels PH (1992) Expert systems for cytometry data analysis. In: Flow cytometry: principles and clinical applications. KE Bauer, RE Duque, TV Shankey, eds. Williams & Wilkins (in press).Google Scholar
- Spath H (1985) Cluster dissection and analysis. Theory, FORTRAN programs and examples. Halsted Press. John Wiley & Sons, New York.Google Scholar