Resources
All tutorials and reference material are available both below and in the MUnitQuest Tutorials repository on GitHub. As registration and submission are handled through Codabench, dedicated platform tutorials are included alongside the technical content below.
Codabench
Codabench is an open-source platform for data science competitions and is used in MUnitQuest to handle submissions to both the Data challenge and the Algorithm challenge. We provide a Codabench Tutorial to help you learn how to register on Codabench, form a Team, make a submission, etc.
Go to the Codabench Tutorial →
Data Challenge
We provide tutorials and tools to guide you through building your first EMG-BIDS dataset.
MUnitQuest EMG-BIDS Tutorial
A step-by-step guide covering how to prepare an HD-EMG dataset in EMG-BIDS format and organize your labeled spike trains (Steps 1 and 2 of the submission process). We developed a simplified approach tailored to typical HD-EMG recordings, walking through five simple CSV template files that generate a complete set of BIDS metadata files, with Python and MATLAB code throughout.
MUnitQuest EMG-BIDS Metadata Tool
An online form that takes your five filled-in CSV files and produces a metadata.zip containing all BIDS sidecar files, ready to accompany your recordings.
MUniverse EMG-BIDS Tutorial (Python)
The Python-based MUniverse package includes a set of classes for handling EMG-BIDS datasets. For example, to read, write, and validate a dataset/recording. We provide a tutorial illustrating how to use these utilities to generate your own EMG-BIDS dataset, including motor unit spike labels.
Open the Jupyter Notebook Tutorial →
BIDS documentation, Examples and Tools
For a deep dive into EMG-BIDS, check out the BIDS documentation and the BIDS examples.
Examples of publicly available EMG-BIDS datasets can be found, e.g., in the MUniverse data collection or on NEMAR.
The online BIDS validator allows you to verify your own BIDS dataset. A list of tools to handle BIDS datasets can be found here.
Algorithm Challenge
Competition Data
The competition data is availible on DaRUS guarnteeing data sharing under the FAIR (Findable, Accesible, Interoperable and Reusable) principles.
Go to the Isometric Familiarization Data Collection →
Go to the Dynamic Familiarization Data Collection →
Tutorials
Jupyter notebooks to help algorithm developers get started with the MUnitQuest test data and submission pipeline.