University of Twente Student Theses

Login
As of Friday, 8 August 2025, the current Student Theses repository is no longer available for thesis uploads. A new Student Theses repository will be available starting Friday, 15 August 2025.

Machine learning for intent recognition in a powered knee prosthesis

Nies, S.H.G. (2020) Machine learning for intent recognition in a powered knee prosthesis.

[img] PDF
1MB
Abstract:To develop a support vector machine (SVM) to predict locomotion mode transitions between level-ground walking and stair ascent, and to investigate the addi-tional value of subject-specific SVMs. Methods: Training data consisted of steady-state level-ground walking and stair ascent, and transitions between them. The effect of input data on the accuracy of the SVM has been analyzed with regard to three aspects: the number of included subjects, the analyzed portion of the gait cycle, and the number of classes. Results: Optimal characteristics of the SVM input data, such as the selected portion of the gait cycle, were subject-specific. The number of classes did not strongly affect the accuracy of the prediction. Conclu-sion: Subjects exhibited slightly different locomotion patterns, affecting the optimal timing of the prediction. Furthermore, a subject-specific SVM resulted in the highest accuracy of intent recognition. Separating transitional and steady-state data into two classes hardly improved the intent recognition accuracy. Significance: This study suggests that the focus should be on subject-specific models for intent recognition, rather than general models that can be applied to all amputees.
Item Type:Essay (Master)
Faculty:TNW: Science and Technology
Subject:30 exact sciences in general
Programme:Biomedical Engineering MSc (66226)
Link to this item:https://purl.utwente.nl/essays/81262
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page