Thanks for the excellent outline!

I think this is an important contribution, indeed the real source of Alpha is not part of LORETA’s solution space which is important to be aware of.

Furthermore, I agree that most research and databases have maybe not looked at alpha the way they should. Individualizing alpha is in my opinion very important. However, this is very hard to automate. Auto-scoring of the iAPF is really hard to do given that there are multiple alpha like sources e.g. central Mu rhythm, often 1-2 Hz faster; parietal alpha, sometimes occipital alpha which is faster tuned than the parietal alpha; frontal alpha which as you stated as well is often slower tuned and other variants such as temporal alpha and tau rhythm when using MEG. I think the only reliably way to score the iAPF is manually. This intrinsically makes it hard for researchers and databases to really incorporate personalized frequency bands based on APF.

Regarding coherence… well… what is coherence….? I think there is not enough standardization in the research and QEEG field to clinically use coherence. The first step is to have guidelines about which methods of coherence to use and understand the implications of for example establishing coherence deviations using QEEG hardware and database X and feeding back coherence with equipment Y… these might both use different expressions for calculating coherence hence implicating that you might feedback something completely different then measured in the QEEG.

Rob Coben and I are trying to get more insight into this by running the same EEG data from a group of Dyslexics through all databases and checking what the differences are. I hope this sheds more light onto this topic…

All the best!

Kind regards,

Martijn Arns
Director / QEEG-D

Brainclinics Diagnostics B.V.
Brainclinics Treatment B.V.
Bijleveldsingel 34
6524 AD Nijmegen
The Netherlands

Tel: +31(0)24-7503505
GSM: +31(0)6-48177919
Fax: +31(0)24-8901447

E-mail: martijn@brainclinics.com
URL: http://www.brainclinics.com