This repository contains the work conducted as part of a university course project focusing on the gait analysis of post-stroke subjects (and one healthy subject).
The research is based on data derived from the paper available at PubMed.
The analysis includes the examination of joint angles and other kinematic parameters (stride speed, length, and duration) as well as the study of muscle activations/EMG.
The study was computed using MATLAB. The code is provided in the scripts folder of this repository. It includes scripts for data processing, analysis, and visualization related to the project's aims.
This repository serves as a comprehensive resource for researchers and students interested in the biomechanics of post-stroke gait rehabilitation, offering insights into the kinematic and kinetic aspects of walking impairments and their potential treatments.
The workflow of this project is outlined as follows:
- Identification of Gait Cycles: Determining the limits of the gait cycles by identifying the minimum of the RHEE marker, located on the right heel, along the vertical axis. A gait cycle starts with an heel strike and ends with the following heel strike.
- Calculation of Reference Systems: Calculating the reference systems of the lower limb segments defined in Vicon protocol.
- Computation of Joint Angles: Calculating the Euler Angles of the lower limb segments.
- Segmentation of Joint Angles into Gait Cycles and Calculation of Mean and Variability: Segmenting the joint angles data into individual gait cycles and computing their mean and variability, using standard deviation and min-max range.
- Calculation of Kinematic Parameters (Speed, Step Length, and Duration): Computing the kinematic parameters to assess the dynamics of the gait cycle further.
- Evaluation of EMG Signals: Evaluating EMG signals quality observing raw signal and periodogram.
- Signal Filtering: Filtering EMG signal to remove noise and artifacts.
- Computation of Signal Envelope: Calculating the envelope of the signals to highlight the overall muscle activation trends.
- Identification of Noise-Only Zones in EMG Signals: Detecting portions of the signal that contain only noise.
- Calculation of Threshold for Muscle Activation Intervals: Establishing a threshold on the signal envelope to accurately identify intervals of muscle activation.
Alongside the code, the project also involved the creation of a detailed report/thesis. However, due to academic requirements, this document was written in Italian rather than English. This may limit accessibility for non-Italian speakers, but it provides comprehensive insights and findings of the research conducted.
This project makes use of chordPiG.m, magnitude.m, makecolumn.m, and makeunit.m functions of the biomechZoo toolbox from the following work:
Dixon PC, Loh JJ, Michaud-Paquette Y, Pearsall DJ. biomechZoo: An open-source toolbox for the processing, analysis, and visualization of biomechanical movement data. Computer Methods and Programs in Biomedicine, Volume 140, 2017, Pages 1-10, https://doi.org/10.1016/j.cmpb.2016.11.007.