This proposal is for a renewal of a project for which the long term goal is to advance the state of the art in the automated analysis and recognition of leukocytes, by computer assisted, high resolution image analysis. As indicated by the minor change in title, the emphasis has shifted from """"""""recognition"""""""" to """"""""analysis"""""""". Thus, the proposed continuation project will concentrate on developing automated techniques for routine measuring and interpreting morphologic characteristics of leukocytes, to supplement present manual and automated analysis, which is limited almost entirely to an enumeration of the relative numbers of the various defined leukocyte types (differential count). Such global measurements of leukocyte morphology (e.g. parameters such as slide-averaged neutrophil nuclear area, or cytoplasm color) are expected to provide additional diagnostic and prognostic information to that contained in the differential count. Routine use of this information has the potential to improve diagnosis and patient management for anumber of diseases and conditions. And since the classification of individual cells requires automated differential counters routinely to make the cell-by-cell measurements from which such averages can be calculated, no new technology need be developed to take advantage of this information. Therefore, a primary thrust of this proposal is to determine which morphologic measurements are meaningful, and under what circumstances. In addition, since some of the required measurements may require greater standardization of parameters (e.g. cytoplasm color), we will concentrate on developing techniques which respond to biological variation but minimize response to artifact. The areas where global parameters are expected to provide useful information are presently: a) early detection of blood infection (sepsis), b) discrimination of viral vs bacterial infections, c) detection and grading infarcts, steroid stimulus, malignancies, and hypertensive effects, and d) earlier detection and monitoring of hematologic and other malignancies. Further, apropos of patient management, we propose to extend our techniques from diagnosis and grading to the following of patients with treated disorders.
