Publication: Analyzing variability in loading impact and peak plantar pressure through the classification of static foot regions
Loading...
Date
2024-07-12
Authors
Nuramalina binti Rosli
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Understanding foot biomechanics and preventing lower limb injuries necessitate the precise evaluation of plantar pressure and foot static or dynamic loading patterns. Although several techniques and algorithms for assessing foot pressure distribution have been thoroughly examined in the past studies, there remains a need for improved foot pressure prediction models to enhance diagnostic and treatment plans. The literature still lacks comprehensive comparisons of various classification algorithms in this context, which hinders the development of optimal models for investigating foot biomechanics. This study aims to evaluate the foot pressure distributions across distinct static and dynamic foot regions, assess loading variability impacts on foot pressures across different foot regions and classify foot pressure based on experimental conditions: static, static load, dynamic, and dynamic load. It involves an assessment of various classification algorithms, highlighting their performance and applicability in an experimental study concerning foot pressure measurements under four scenarios: static, static load, dynamic, and dynamic load. The study methodology included data preprocessing, feature selection, classification, and performance evaluation using metrics such as classification accuracy, precision, recall, and F1-score. Six classification algorithms: J48, Random Forest, Support Vector Machine (SVM), Naive Bayes, k-Nearest Neighbors (k-NN), and Logistic Regression algorithms in Weka tool were used to classify plantar pressure measurement. The results showed that the Random Forest and Logistic Regression algorithms outperformed the other models achieving classification accuracies of 58.8275% and 50.2758% respectively. These findings imply that incorporating machine learning classification techniques can significantly enhance the predictive foot plantar pressure. This study paves the way for further advancements in foot biomechanics research, sports performance enhancement and rehabilitation protocols in clinical foot health.