Musclesense workbench is an AI-driven, open-source software for neuromuscular diseases imaging research, that is maintained and developed by a consortium of higher education institutions.
The software offers cross-platform MRI visualisation, processing, and analysis tools all within one platform.
The software was produced as part of the Wellcome Institutional Strategic Support Fund (ISSF3) – AI in Healthcare Call 2019 project “Towards improving the clinical care of patients with neuromuscular diseases using innovative artificial intelligence imaging methods”.
Funded by the Wellcome Trust and the National Institute for Health Research Biomedical Research Centre at University College London and University College London Hospitals NHS Foundation Trust, the project aimed to contribute towards improving the clinical care of patients with neuromuscular diseases using innovative artificial intelligence imaging methods.
Future versions of the software will allow the use of different muscle segmentation tools utilising alternate methods and/or acquisition sequences using a configurable plug-in mechanism. The default deep-learning-based segmentation tool provided with the software is described in the article linked here.
The current consortium members are University College London (UK), and Newcastle University (UK). Please email us to express interest in joining the consortium.
Download the latest release version. Alternatively, you can download or clone the repository instead if you would like to test the latest development version.
Simply save and extract your download at a location of your choice on your computer.
It is no longer necessary to download any model weights as these are automatically retrieved by the software as required.
The following are prerequisites (where appropriate with example commands):
itk-snap
Python3
sudo apt install python3-pip
sudo apt install python3-tk
python3 -m pip install numpy --user
python3 -m pip install nibabel --user
python3 -m pip install matplotlib --user
python3 -m pip install pandas --user
python3 -m pip install joblib --user
python3 -m pip install scikit-learn --user
python3 -m pip install tensorflow==2.2.0 --user
python3 -m pip install keras==2.3.1 --user
python3 -m pip install h5py==2.10.0 --user
python3 -m pip install scikit-image --user
python3 -m pip install pillow --user --upgrade
python3 setup.py (from the installation directory)
To run the workbench, type python3 mmseg_app.py from the installation directory
While the above example commands are given for installation on an Ubuntu OS, the software should work on other Linux OSes as well as MacOS. It will probably need some modifications before it will work on Windows though. Do not hesitate to get in touch if you have any trouble installing and getting the software up-and-running on your system.
Musclesense the algorithm, and Musclesense Workbench should not be used in the diagnosis or treatment of patients.
Please submit any enquiries here