[Eeglablist] Pre-Processing Continuous EEG study using that used two different devices (BDF/EDF)

[Brian Harvey]

Hi all,

I am tasked with the analyses of EEG acquired as part of a natural history study for a pediatric patient population. These EEGs were acquired using two generations of Lifelines Neuro ambulatory EEG amplifiers (Mk3 and T4A). The Mk3 device is AC-coupled and uses a single ground/reference (central/midline) which saves raw data as EDF (16 bit ) while the other (T4A) uses a ground and CMS/DRL reference scheme (also midline/central) and saves raw data as BDF (24bit resolution). I am familiar with Biosemi hardware (active two and pre-clinical activeRat version) and realize that without choosing a channel for reference at loading will result in a less than desired SNR/CMRR.

I am open to anyone that can provide insight/suggestions for how to re-reference these data such that the result is as congruent as possible. What are the implications of such as mixed acquisition when pooling the entire dataset after any pre-processing steps?

Thanks in advance,


Brian Harvey

Neural Circuits

Biogen Inc.

brian.harvey at biogen.com