Feature vectors of the cruciate ligaments of the knee joint

Abstract

The most important and primary aims of this article are two elements. The first one is a presentation of the feature vectors of the anterior and posterior cruciate ligaments of the knee joint and the second element is a discussion on the choice of the most effective features. In order to build the feature vectors, the extraction of the cruciate ligaments structures from the MRI images of the knee joint was necessary. This operation has been made on the basis of the following fuzzy methods: fuzzy C-means algorithm with median modification (in order to find a region of interest including cruciate ligaments), and fuzzy connectedness (in order to extract the cruciate ligaments). The presented methodology has been tested on 74 clinical T1-weighted MRI slices of the knee joint. On the basis of the described methodology a software application has been built. This software application is dedicated for the anterior and posterior cruciate ligaments diagnostics and it seems to be a very helpful for the orthopedists.