Automation in cervical cancer screening. Part 1: fixed cell scanning systems

Cytology automation is highly desirable if a mass cancer screening system is to be effective, and has been the subject of widespread research and development effort. Several experimental systems have been developed based on fixed cell scanning or flow techniques, but clinical trials have produced disappointing results. The major problem with these early systems has been that of false positive "abnormal cell" signals caused by cell overlaps and artifacts. Most current research is devoted to the development of better techniques for cell presentation, to the improvement of pattern recognition techniques, and to the exploitation of novel cell parameters in flow systems.