Assessment and classification of lower back pain

Abstract

This thesis presents work done towards developing a reliable determining system to assess and classify the lower back pain. In the initial stage, we seek to explore the possibilities of coming up an effective and user-friendly way to assess patients with non-specific lower back pain thereby ValedoTM motion system. However, the data collected by ValedoTM motion sensors cannot be directly exported for in-depth analysis due to the commercial confidentiality. Therefore, another novel determining system has proposed to assess the lower back pain patients based on Maximum-Likelihood Estimation of Gaussian Mixture Model algorithm. To validate the proposed classification algorithm, a motion testing experiment is conducted to analyze the motion data from healthy participants and LBP patients by collecting the data from DelsysTM TrignoTM EMG system.