Development and application of an open-source processing pipeline for magnetic resonance imaging (MRI) of muscle in muscular dystrophies

Background

Muscular dystrophies are muscle disorders where progressive muscle damage leads to replacement of muscle tissue with fibrosis and fat, with patients ultimately losing strength and independence. Currently, there are no treatments that significantly slow progression of these diseases, but several pharmaceuticals are in development. Magnetic resonance imaging (MRI) is a non-invasive, non-ionising technique that has proved invaluable in understanding disease progression in muscular dystrophies. Our currently-running D-TERMINED study is using MRI to measure muscle properties such as fibre size and membrane permeability – which are implicated in the early phases of muscular dystrophies. We hope these measures will be used for identifying a suitable window for treatment and for monitoring muscle changes during pharmaceutical trials. However, currently, processing of the MRI data is a complicated, multi-step process, relying on MATLAB code in tandem with third-party executable applications. We are seeking an enthusiastic student to assist with converting the existing code into a streamlined pipeline, written in Python. This code will then be applied to analyse the existing MRI data, and will ultimately be shared with the scientific community as an open source package.

Description of the Project

This internship is attached to the D-TERMINED project, which is part of a five-year natural history study involving up to 50 genetically confirmed Becker muscular dystrophy (BMD) patients, and 20 age-matched volunteers. The aim of D-TERMINED is to test diffusion-tensor MRI (see figure) as a new ‘imaging biopsy’ for studying muscle changes over time and for evaluating new treatments. Each patient has undergone MRI twice, with visits being one year apart. Healthy volunteers undergo MRI either once or twice – the former undergoing a muscle biopsy, and the latter being assessed for short-term MRI repeatability. With this study design, at least 70 complex MRI datasets will be available. This large volume of scans is challenging to process and analyse, but will give import insights into disease progression in BMD.

Internship goal and structure

• Get familiar with the existing MATLAB pipeline and MRI data
• In parallel:
  • Begin translating the MATLAB pipeline to Python, evaluating as you go that performance is maintained and outputs are comparable;
  • Process the available MRI data to produce maps of muscle fibre size and membrane permeability.
  • If time permits, perform muscle segmentation and extract quantitative metrics (fibre size and membrane permeability) for multiple muscles per patient.

Required experience

• Coding experience in Python and/or MATLAB
• Some background in image processing / analysis preferred

Application procedure

• If you are interested, please send a short motivation letter and your CV to: Donnie Cameron, PhD: Donnie.Cameron@radboudumc.nl

Please feel free to reach out with any questions regarding the content of the internship.