Research Project
3508360
Phase I will explore, via computer simulation, in a very high-speed computer the use of state-of-the-art RCA Charge Coupled Devices (CCD) in an Automated Diagnostic Microscope (ADM) for imaging and analyzing blood cells stained with a quantitative fluorescent technique that allows accurate characterization of all of the types of cells encountered in periferal blood. Digital images prepared from calibrated microphotographs of white blood cells, platelets, and red cells, will be used to represent the output of the CCD imagers in the computer simulation. The simulation will then be used to design the prototype instrument. Phase I will also include the structuring of the simulated image data into parameters related to the quantity and morphology of cell constituents (DNA, nuclear size and shape, heterochromatin, cytoplasmic volume and shape, hemoglobin, specific granules, lysosomes, and RNA). Software will then be developed and explored for using these parameters to classify and describe blood cells. The parameters will then be used to guide the development of cell analysis software. The technique of computer simulation should very significantly aid in speedy production of a functioning prototype ADM in Phase II. The ADM will represent a significant advance in blood cell imaging because of its high resolution and sensitivity that capitalize on a quantitatively interpretable specimen. The ADM will expand the data base describing normal and abnormal (including neoplastic) blood and bone marrow cells, providing quantitative diagnostic parameters that are not currently available. The first level design goal is, however, to provide all of the currently available hematologic tests made on blood cells in a completely automated mode. The ADM will be extremely cost-effective. It will perform all the tests done on blood cells in the clinical hematology laboratory at least as fast and in more detail than current instrumentation. The ADM will be capable of the diagnostic discrimination of blood cells, and will be easily adapted to the utilization of monoclonal antibodies in cell identification and characterization. These factors should stimulate commercial interest.
