Heat map visualization of multi-slice medical images through correspondence matching of video frames

摘要

Diagnosis by radiographers through examination of multi-slice medical image data is common, but the process which experts use to arrive at their conclusions is often difficult for novices and medical trainees to understand. We demonstrate a method for capturing and visualizing patterns of eye gaze data for medical professionals viewing such images. We can automatically generate content to quickly identify areas of interest and search patterns used during diagnosis. The data gathered can be used to better understand the visual diagnostic process and aid in creating better automated systems.