Enhancement of Visual Perception Through Dynamic Cues: An Application to Mammograms

Abstract

Images often contain very small and hardly detectable objects or patterns, which can be of grave importance for diagnosis. In this paper we present a new method for aiding medical doctors in diagnosis of such images by adding artificial movement to the static images, in order to utilize the motion sensitivity of the human visual system. This technique permits detection of lesions not just by intensity, structure and texture differences with its surroundings, but by motion as well. Statistical analysis of experimental tests with both radiologists and non-radiologists show improved detection rate of microcalcifications in mammograms, raising it on average by 20.8% for non-radiologists and by 8.4% for radiologists.