Intelligent Orthopedic Diagnosis: A Machine Learning Approach for Biomechanical Data Classification

Abstract

The biggest challenge in the medical research is the detection of orthopedic diseases in an accurate and automated manner. Adding Machine learning in the challenge will create a highlight in the diagnosis. This study proposes a unique model of classification which uses the biomechanical data for the detection of orthopedic diseases. This method achieves a high accuracy of around 97.85%. In our investigations, we perform Exploratory Data Analysis to understand patterns and relationships with regard to the orthopedic data. Recursive Feature Elimination (RFE) selects the significant features contributing to accurate disease detection. SMOTE application is an artificial generation of samples to improve the performance of models, thus addressing class imbalance. The unwanted false positives and the improvement of precision revels that the proposed model has more advantages of traditional methods. Preprocessing the available data and the choice of the model and the performance and the attainment of accuracy offers the solution to the machine learning algorithms in the detection and diagnosis of orthopedic diseases.

Ähnliche Arbeiten