Prediction of Pathologist Navigation Patterns in Virtual Microscopy based on a Soft-Computing Approach

F. Gómez, M. Iregui, and E. Romero (Colombia)


virtual microscopy, user modelling, large images.


Navigation through large digital microscopic images is a potential benefit for histology or pathology teaching, qual ity of diagnosis or communication between pathologists. However, the size of this kind of images is prohibitive for navigation with conventional techniques. This paper presents a navigation model which permits to anticipate the pathologist trajectories tasks when exploring virtual slides. The full strategy combines an offline model of the patholo gist knowledge and a prediction online module which cap tures a particular pathologist navigation pattern. While op timal parameters for the offline model are calculated using an Expectation-Maximizationstrategy, prediction is carried out by a particle filter. Parameters are estimated from sev eral series of actual navigations, pathologists and virtual slides. The present approach is compared with other con ventional prediction methods and decreases the calculated MSE in about a 50 % for the entire group of pathologists.

Important Links:

Go Back