Max Ordinal Learning Applied to the BI-RADS Classification of Mammograms

Authors

Abstract

The main goal of this work is to provide medical experts with a second opinion on breast cancer BI-RADS classification. In order to achieve this goal, we began by performing analysis of an individual images. However, in a real-world scenario, medical experts use all of the available information in order to take a decision. To approximate this situation a breast analysis was performed using two available sights. In future work we intend to expand this work further by gathering ultrasound information. Another future line of research will be to develop adequate feature selection methodologies specific to this particular ordinal application.

Conference

StudECE 2012