Human skeletal muscles are composed of many fibres that, on excitation by the nervous system, contract and produce mechanical force to move the skeleton. The fibres within individual muscles are grouped into distinct functional units, motor units (MUs), and the properties and behaviours of MUs change with age, disease and injury. Electromyography (EMG) is commonly used to study MU properties and behaviours. However, force production relies on excitation and subsequent mechanical contraction of the muscle fibres. EMG alone cannot reveal these mechanical effects, and this represents a gap in the tools available to study this important feature of muscle function.
We have developed some initial ultrasound image analysis tools for assessment of these mechanical effects. The proposed project will therefore harness machine learning image analysis approaches to enhance our initial tool, extending their application and scope and working towards translation to wider research use and clinical practice.
The project is expected to start in January 2023, and the successful candidate will become a part of our Musculoskeletal Science and Sports Medicine Research Centre.
Aims and objectives
The project aim is to enhance ultrasound imaging techniques to improve understanding of skeletal muscle physical properties. The key objectives are:
- Develop a machine learning-based model to automatically extract muscle shape features (fascicle length, orientation, density) from standard frame rate, B-mode ultrasound image sequences, providing useable data for musculoskeletal researchers
- Extend the model to analyse individual motor unit physical characteristics in high frame rate image sequences
- Apply the models to provide preliminary evidence of novel differences in muscle properties between healthy younger and older adults
Full details including how to apply available at: https://www.mmu.ac.uk/research/research-study/scholarships#ai-67266-0