EEG preprocessing/processing using MNE-Python

Background

I am a PhD student at Concordia University enroll in an Individualized program (intersection of cognitive neurosciences and digital arts/electroacoustics) My main PhD focus is on electrophysiological correlates of creativity and pareidolia, as well as how to use this knowledge in Brain-Computer Interfaces (BCI)

But...still waiting for my EEG data!

Instead... Music Imagery Information Retrieval: 10 subjects, 64 EEG Channels for a music imagery task of 12 different pieces w/ different meter, length and tempo

• (https://github.com/sstober/openmiir) - Github
• (https://pdfs.semanticscholar.org/cde4/b1ec89f2c05a41f1143792a890a00e89541a.pdf) - Article

I chose to work with these data so it provides me with a first hands-on experience with EEG data. The produced scripts/notebooks should be reusable for the data I will be collecting in the next few months (hopefully) and for anybody interested in using MNE-Python and complexity measures on electrophysiological data. This project uses open-source EEG data collected on a Music Perception and Imagery task.

Preprocessing

• Install MNE-Python
• Load the data into a Jupyter Notebook
• Visualize raw signal
• Independant Component Analysis (ICA)
• Artifact rejection using ICA
• Epoching (channel x condition x trial)

Processing/Feature extraction

• Compute Power Spectral Density (PSD) averaged by condition
• Complexity measures by channel/song/condition/participant
• Topomaps of complexity measures
• Steady-state-evoked-potential (SS-EP) for each song averaged across subject

• Beat per minute (BPM) for each song

Analyses

• Compare perceived vs. imagined conditions for each frequency band
• Compare PSD topomap between clustered regions
• Compare complexity measures topomap between clustered regions
• Compare complexity measures for each conditions with complexity measures of stimuli (e.g. correlations of stimuli fractality with brain signal fractality)
• Compare SS-EP for binary vs. ternary songs for each conditions
• [ ]

Deliverables

Commented Jupyter Notebook..

• For preprocessing and epoching of data

• For computation of complexity measures, PSD and related topomaps

• Draft of a paper if any promising results!

• Requirements.txt

• Wrap-up on Binder