Prediction model of subacromial pain syndrome in assembly workers using shoulder range of motion and muscle strength based on support vector machine

Abstract

Subacromial pain syndrome (SAPS) is the most common upper-extremity musculoskeletal problem among workers. In this study, a machine learning model was built to predict and classify the presence or absence of SAPS in assembly workers with shoulder joint range of motion (ROM) and muscle strength data using support vector machine (SVM). Permutation importance was used to determine important variables for predicting workers with or without SAPS. The accuracy of the support vector classifier (SVC) polynomial model for classifying workers with SAPS was 82.4%. The important variables in model construction were internal rotation and abduction of shoulder ROM and internal rotation of shoulder muscle strength. It is possible to accurately perform SAPS classification of workers with relatively easy-to-obtain shoulder ROM and muscle strength data using this model. In addition, preventing SAPS in workers is possible by adjusting the factors affecting model building using exercise or rehabilitation programs. Practitioner summary: This study aimed to create a machine learning model that can predict and classify SAPS using shoulder ROM and muscle strength and identify the variables that are of high importance in model construction. This model could be used to predict or classify workers’ SAPS and manage or prevent SAPS.