In this research, the primary goal is to develop an automated cervical smear analysis system for use after manual screening of thinlayer cervical specimens as a quality control (QC) instrument. The device will measure nuclear features termed """"""""Malignancy Associated Changes"""""""" (MACs) to identify disease in samples that are falsely negative by manual methods. At the current levels of technology and in the present regulatory environment, the development of such a system will play an important role in laboratory quality, and will likely reach the market long before an automated primary screening device. The use of MACs is new and innovative in that it allows fast and accurate processing at a very reasonable system cost. In Phase II a fully automated prototype would be developed and tested. Specific research areas are 1) the optimization of the stain, 2) the use of multiple focal planes, 3) optimization of classifiers, and 4) completion and testing of a completely automated prototype. Specification and design of a commercial instrument will begin at the end of Phase II and continue into Phase III.
This device would provide enhanced quality control for Pap test analysis in all cytology laboratories worldwide. The current practice of rescreening 10% of negative smears is not adequate. Implementation of a fast and inexpensive automated system used after manual screening would result in increased accuracy and improved healthcare.
