ID 2453: Monitoring prodromal hyperkinetic movements using body worn sensors in Huntington’s Disease
Huntington’s disease (HD) is a progressive neurodegenerative disorder that affects movement, cognition, and emotion. It is characterized by the gradual degeneration of neurons in the brain, particularly in regions responsible for motor control and coordination. The disease often begins in the prodromal phase, where individuals may experience subtle changes such as clumsiness, altered gait, or difficulties with fine motor tasks, alongside potential mood or cognitive disturbances. As the disease progresses to the manifest phase, these early motor dysfunctions develop into more pronounced symptoms, including chorea (involuntary jerky movements), muscle rigidity, and impaired balance and coordination. This transition reflects the increasing neurological damage and marks the onset of significant motor disability, severely impacting daily functioning and independence.
This project aims to develop AI-based algorithms using data from body-worn sensors to predict the transition from premanifest to prodromal and motor Huntington’s Disease (HD). We will analyze sensor signals to identify features that distinguish prodromal and early manifest HD from premanifest stages. Sensors placed on the feet, wrists, and trunk will be used to detect movement abnormalities, with a particular focus on hyperkinetic movements such as fidgeting and chorea.
Dataset
The dataset consist of Premanifest, Manifest and control subjects. Each subject wore with five sensors on the ankles, wrists and lower back while performing five different tasks of sitting, pouring a glass of water and drinking, watching TV, serial seven tasks, and explaining a picture. The Primary research question is the classification of these three groups while comparing different sensor placements and tasks.
Requirements
- Proficiency in Python programming language
- Proficiency in basic libraries such as Pandas and Numpy, and Scikit-learn.
- Understanding of signal processing and feature extraction.
- Familiarity with various types, concepts, implementations, and testing of machine learning (ML) algorithms.
- Great scientific writing skills.
- Communication skills (English).
Tasks
- Literature search.
- Data cleaning and pre-processing.
- Manual Label checking. The labeling has been done, but checking the labels might be required.
- Feature extraction, including features extracted using CNN.
- Implementing different ML algorithms for classification
- comparison of the different sensor placement and tasks.
- Writing.
Supervisor:
Please use the application form to apply for the topic. We will then get in contact with you.