I need to preprocess EEG data from 350 ICU patients, all of whom are comatose. Each patient has 72 EEG files, each containing 1 hour of recording. The files are in MATLAB format (.mat).
I’ve started working on the preprocessing, but I’m feeling really stuck and overwhelmed. I’m unsure whether I need to manually perform steps like ICA or what the complete preprocessing pipeline should look like. Also, the EEG recordings contain many flat lines and artifacts, which makes it even more confusing how to proceed.
Since I have to complete this within three months, I’m extremely worried about where to start and how to handle such a massive dataset properly and efficiently.
Could anyone kindly suggest:
Whether ICA and other preprocessing steps should be done manually or if there are good automated options?
How to deal with flat-line EEG segments — should they be removed before ICA or handled differently?
Any good resources (tutorials, papers, lectures) to properly understand EEG preprocessing, especially for ICU/comatose patients?
Recommended preprocessing pipelines or toolboxes that can handle large datasets?
I would be immensely grateful for any advice, references, or guidance. I really need urgent help to get started correctly.