Image Processing in Medical Microscopy

Abstract

Full automation in medical microscopy has been accomplished in the field of clinical determination of the white blood cell differential count. Manufacture of differential counting microscopes commenced in 1974, and approximately 1,000 of these robots are now in the field. They analyze images of human white blood cells, red blood cells, and platelets at the global rate of approximately 100,000 slides per day. This incredible throughout represents automated image analysis and pattern recognition at the rate of 5 billion images per year and represents a major accomplishments in the application of machine vision in medicine. In other areas, such as cytology and cytogenetics, automated computer vision is still in the research phase. This paper discusses the state of the art in blood smear analysis automation and in other related areas including multi-resolution microscopy where images are currently being generated over a 64:1 magnification containing from one-quarter megapixel to one gigapixel in full color.