Ensemble-Based Orthopedic Biomechanical Characteristics Prediction Using Probabilistic Reasoning And Machine Learning Approach

Abstract

In hospitals, medical data are growing increasingly complicated and heterogeneous in significant data circumstances. Massive medical data management requirements have yet to be satisfactorily addressed by the conventional way of manual computation. With the advancement of artificial intelligence and machine learning, this medical diagnosis model has been created. Now withstanding to enhance the perforation in the additional determination of orthopedic disease data. The purpose of this paper is to provide a machine-learning technique for auxiliary categorization prediction that may be used to aid in the diagnosis of orthopedic diseases. This study shows how to build ensembles of diverse classifiers by stacking many types of classifiers,

including the Gaussian NB, logistic regression, decision tree regressor, K-Nearest Neighbor algorithm (KNN), and Support Vector Machines (SVM). Five different base classifiers are used, and min-max ranking is used to give weightage to various base classifiers. This work is crucial because the proposed algorithms can make quick judgments on diagnoses of orthopedics with a good level of accuracy. The ensemble stacking, the ensemble stacking with equal weightage, and the ensemble stacking giving rank by utilizing the min-max method are all proposed in this study. 

• The findings of this research indicate that stacking, followed by ranking, using the min-max method can be used in developing expert systems that are both efficient and effective in diagnosing disc hernia and spondylolisthesis.


• This study suggests an orthopedic diagnosis classification and prediction model with an accuracy of 97.80% and 97.85%.


• The proposed system aims to reduce medical staff labor, helps patients prevent and recover early, and provides real supplemental clinical attention.