Neuroinformatics
I build open-source tools for EEG and neuroimaging research. These tools are used by thousands of researchers worldwide and have collectively been cited tens of thousands of times.
NEMAR
NEMAR (Neuroelectromagnetic Data Archive and Tools Resource) is an open archive for sharing and analyzing EEG, MEG, and iEEG data. Built on the OpenNeuro platform, NEMAR provides standardized BIDS-formatted datasets with integrated processing pipelines.
EEGLAB
EEGLAB is an open-source MATLAB/Octave toolbox for processing and analyzing EEG, MEG, and other electrophysiological data. I created EEGLAB in 2000 during my post-doc at the Salk Institute and partnered with Scott Makeig to expand it into a comprehensive platform incorporating Independent Component Analysis and time-frequency analysis.
EEGLAB is the most widely used EEG analysis software, with over 47,000 citations of the reference article. It supports a plugin architecture with dozens of extensions for source localization, connectivity analysis, artifact rejection, and more.
EEGPrep is a Python implementation of the EEGLAB preprocessing pipeline, bringing EEGLAB's methods into the Python ecosystem. [GitHub]
I also wrote ICA for Dummies, a tutorial explaining Independent Component Analysis in plain language for researchers new to the method.
EEGDash
EEGDash is a deep learning interface to the NEMAR archive, enabling AI-driven exploration and analysis of open neuroimaging data. It provides automated feature extraction and classification tools for large-scale EEG datasets.
SpikeNET
During my PhD in Toulouse, I developed SpikeNet -- a neural network simulator using asynchronous spike-based processing for ultra-rapid visual recognition. This work led to co-founding SpikeNet Technology in 1999, a startup commercializing the technology for real-time face and object detection. The technology was later acquired by BrainChip Holdings, a neuromorphic computing company. Read more about SpikeNET and the origin of convolutional neural networks.