| Summary: | Knee Osteoarthritis (KOA) is a multifactorial disease that causes low quality of life, poor psychology and resignation from life. Furthermore, KOA is a big data problem in terms of data complexity, heterogeneity and size as it has been commonly considered in the literature with most of the reported studies being limited in the amount of information they can adequately process. The aim of this paper is: (i) To provide a robust feature selection (FS) approach that could identify important risk factors which contribute to the prediction of KOA and (ii) to develop machine learning (ML) prediction models for KOA. The current study considers multidisciplinary data from the osteoarthritis initiative (OAI) database, the available features of which come from heterogeneous sources such as questionnaire data, physical activity indexes, self-reported data about joint symptoms, disability and function as well as general health and physical exams’ data. The novelty of the proposed FS methodology lies on the combination of different well-known approaches including filter, wrapper and embedded techniques, whereas feature ranking is decided on the basis of a majority vote scheme to avoid bias. The validation of the selected factors was performed in data subgroups employing seven well-known classifiers in five different approaches. A 74.07% classification accuracy was achieved by SVM on the group of the first fifty-five selected risk factors. The effectiveness of the proposed approach was evaluated in a comparative analysis with respect to classification errors and confusion matrices to confirm its clinical relevance. The results are the basis for the development of reliable tools for the prediction of KOA progression.
|
|---|
