Knee injury multi-label classification with label dependencies

Sport knee injuries are the leading causes for most knee surgeries performed annually. Anterior cruciate ligament (ACL) tears and Meniscus tears are the most prevalent injuries to occur among people and athletes. The injuries are often detected using arthroscopy or knee magnetic resonance imaging (MRI). Arthroscopy is considered as an invasive method to analyze knee injuries; therefore, knee MRIs are more preferred for diagnosis.

However, analyzing the MRIs is considered time consuming for the radiologists and it might be affected by different human errors and interpolations. Hence, it is important to have an automatic tool to read these MRIs and classify the injury type and intensity level. As a result, machine learning algorithms have been used to develop automatic classifiers to classify the different injuries present within the medical images. In this project, a machine learning algorithm is developed to classify the different Knee injuries presented in the knee MRIs utilizing the possible dependencies between the injuries.