نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسندگان English
Falling is a major public health concern, especially for those dealing with low back pain, as it can lead to severe physical and psychological setbacks. While traditional clinical tests often miss the subtle nuances of balance impairment, quantitative Center of Pressure metrics offer a much clearer picture of stability. By combining these COP features with the right exercise intensity and advanced machine learning, we can create a more powerful tool for predicting and analyzing balance issues in patients with back pain.
In this study, we collected Center of Pressure data and extracted the key features defining postural stability. Following rigorous preprocessing and normalization, we evaluated several machine learning models specifically Logistic Regression, Random Forest, and Support Vector Machines to classify individuals with stable versus impaired balance.
The results demonstrated that these machine learning architectures performed exceptionally well in distinguishing between the two groups. Among the features extracted from the Center of Pressure data, velocity metrics and displacement amplitude emerged as the most potent predictors of poor balance. Through stratified cross-validation, our models achieved high accuracy and stability, providing an effective and reliable framework for identifying balance deficits
کلیدواژهها English
