Neurosciences on Mathieu Scheltienne

NiGSP

Mon, 01 Apr 2024 00:00:00 +0000

NiGSP is a python library to run Graph Signal Processing on multimodal MRI data, developed by the Medical Image Processing Lab (MIP:Lab) headed by Prof. Dimitri Van De Ville.

After working for 2 years in this amazing lab, I contributed to NiGSP to improve the unit testing, CI infrastructure, documentation build and overall code quality. The project was originally developed by Stefano Moia (github handle: @smoia).

MNE-LSL

Sun, 01 Oct 2023 00:00:00 +0000

MNE-LSL provides a real-time brain signal streaming framework tightly integrated with MNE-Python. It provides a new Python binding for the Lab Streaming Layer C++ library, an interface widely adopted by EEG manufacturers to stream acquisition data in real-time.

I took over NeuroDecode, a project initially developed by the Fondation Campus Biotech Geneva and transformed it into software benefiting from the robust and well-used MNE-Python API. The project was rebranded into MNE-LSL and joined the MNE ecosystem in October 2023.

MNE-Python

Wed, 01 Mar 2023 00:00:00 +0000

MNE-Python is one of the 4 major M/EEG analysis packages, alongside its 3 MATLAB counterparts, FieldTrip, Brainstorm and EEGLAB. After using it and contributing for several years, I joined the maintainers in March 2023.

Pycrostates

Wed, 01 Jun 2022 00:00:00 +0000

Pycrostates provides EEG microstates analysis tightly integrated with MNE-Python and opens the door for microstates analysis applied to other sensor types, for instance MEG magnetometers and gradiometers.

The project was developed in collaboration with Victor Férat (github handle: @vferat).

MNE-ICALabel

Sun, 01 May 2022 00:00:00 +0000

Scalp electroencephalography (EEG) and magnetoencephalography (MEG) analysis is typically very noisy and contains various non-neural signals, such as heartbeat artifacts. Independent Component Analysis (ICA) is a common procedure to remove these artifacts. However, removing artifacts requires manual annotation of ICA components, which is subject to human error and very laborious when operating on large datasets. The first versions of mne-icalabel replicated the popular ICLabel model for Python.

The project was developed in collaboration with Adam Li (github handle: @adam2392) and Jacob Feitelberg (github handle: @jacobf18).