An Original E-Testing Method for Medical Learning

Abstract

The paper presents an original and complex e-testing modality, which adds to the traditional methods used in medical learning. The new proposed modality implies the existence of a database with medical images and uses a modern querying method: content-based visual query. The testing method resides in choosing an image/region as query for which the system will return a number of images which are visually similar and could be relevant or non-relevant (belonging or not to the same diagnosis class). The student has to recognize the diagnosis and mark the relevant images. The option of visualizing the correct solution to the test helps the student to see what errors he made. This method offers the possibility to test the ability of the student in recognizing images that belong to the same diagnosis or which are similar but are associated to a different diagnosis. The teacher can determine the level of knowledge achieved by the student regarding a certain pathology and also in general, as well as the studentpsilas capacity to establish a correct diagnosis based on imagistic investigations frequently used in todaypsilas medical practice.