Courses

Alex Hughes

BSc [Hons] Creative Computing

Hey I'm Alex, I'm a final year Creative Computing student in IADT! 👋 I'm super passionate about creating applications and participating in projects with the goal of increasing the quality of life in various industries, recently medical technology! This passion has been the driving force behind my latest project which is Neural Signal Interpretation for Prosthetics. My thesis project aims to create a muscle signal acquisition and processing system that can be built both from cheaper components and with less overall complexity than regular industry prosthesis, while also retaining the same quality of functionality! 🦾

In the beginning of my project I used a dataset created by Mendeley Data to use sampled EMG signals to design the foundation functionality that would carry me through to the rest of the project. The dataset comprises raw EMG data collected via sensors attached to the muscle of interest. This data is then meticulously processed using various digital signal processing techniques to enhance its clarity and usability. Key steps in the processing pipeline include bandpass filtering to remove noise outside the muscle activity frequency range, rectification to ensure the data reflects the absolute signal strength, and smoothing to highlight the overall muscle activation trends. These steps transform the raw, noisy EMG signals into clear indicators of muscle activity intensity. In the video demonstration included, I showcase the practical application of this dataset through an Arduino setup. The Arduino is programmed to interpret processed EMG signals in real-time and adjust the colours of an RGB LED to reflect the intensity of muscle contractions. Lower signal strengths trigger a red light, moderate strengths switch the light to green, and high strengths turn the light blue. This colour-coded feedback provides an immediate visual representation of the muscle activity. The demonstration highlights the direct application in creating interactive and responsive hardware interfaces.

There are many stages I go through in the filtering/processing stages of the datasets initial signals, these are filtering using a bandpass, rectifying the signals to isolate the positive values and enveloping to get the smooth average values which can be read easily during feature extraction or gesture recognition.

Here I showcase the intricate setup and functionality of the components connected to the Grove board in this video, a crucial aspect of my project dedicated to the acquisition and processing of live electromyography (EMG) signals. The Grove board, known for its modular and user-friendly interface, serves as the central hub for connecting various sensors and modules that play pivotal roles in capturing and initial processing of the live EMG data. The setup includes an EMG sensor (electrodes) which can be directly attached to the muscle, which detects electrical activity generated during muscle contractions. This sensor outputs analogue signals that correspond to the strength and frequency of muscle activity. These signals are then fed into the Grove board, which interfaces with additional modules for enhanced signal processing. In the video demonstration, you can see the following key components: EMG electrodes. Amplifier Module (Grove Shield). Grove board (Arduino) Breadboard with LEDs for testing. USB for power and data output. All these components combined make up the hardware necessary to acquire EMG signals to then be sent to processing and received again for either testing with the LEDs or another output.

Once the live data is acquired and processed, I then use these signals to control an array of 5 LEDs based on the signal intensity. This intensity is defined using thresholds to accurately convey how much power the electrodes are receiving which goes on to send actionable commands to my circuit from the Arduino. This showcases the real-time application of my research and development with the outcome of communicating effectively with real world components being controlled by the live signals.

When processing the signals from the electrodes, there are a few more steps than with the dataset. The differences is that there are 3 different processing/filtering stages the raw EMG data must pass through first. These are a Bandpass filter, a Butterworth filter and finally a Discrete Wavelet Transform filter. These 3 combined are used to process the raw data which can then be sent away for the classification and feature extraction stage.

Neural Signal Interpretation for Prosthetics

My project aims to tackle the persistent challenges within the prosthetic field of signal interference, high cost and component complexity by simplifying the signal acquisition and processing stages, ultimately making prosthetic technology both more affordable and functional. Through detailed research into both Electromyography (EMG) which is the study of signals collected from the muscles, the development of tailored designs, and the application of novel processing techniques, my work will contribute to the broader goal of democratizing access to advanced prosthetic devices.

View My Thesis Here! DOCX

Neural Signal Interpretation for Prosthetics

My thesis explores innovative methods to simplify the signal acquisition and processing stages of EMG systems, aiming to reduce costs and enhance accessibility for a broader population. Through a comprehensive study integrating both Electroencephalogram (EEG) and EMG techniques, my research employs non-invasive methods to optimize signal fidelity and user comfort.

The work includes the development of a prototype using Arduino and Grove systems for real-time signal processing, tested against a dataset provided by Mendeley Data. The findings suggest that simplifying these systems without sacrificing functionality is feasible, paving the way for more affordable and accessible prosthetic solutions.

The thesis contributes to the field of biomedical engineering by demonstrating how cutting-edge technology can be adapted to meet the needs of diverse users, potentially enhancing the quality of life for individuals with limb loss.

View My Thesis Here! DOCX

Creative Computing

2021 2022 2023 2024

All coursesClose

Top