Dear all,
when new users arrive at YNiC, some would like to practice using data
analysis scripts before they record their own data. As such, we would
like to document some of the paradigms that have already been recorded.
We would not want to know in detail the specifics of your project, just
the basics to allow for a sensible analysis.
For example, the following may be enough:
"In the data for project Pxxxx, responses were recorded in continuous
mode to visual stimuli. There were four conditions which had the
respective trigger codes 4150, 4152, 4154 and 4156. Triggers 4150 and
4152 had a group code of 2, and triggers 4154 and 4156 had group codes
of 4. Participants made responses to some trials, which have a response
code of 32. Suggested epochs for data analysis are 150ms pre-trigger and
900ms post-trigger. Data was recorded for participants Rxxxx, Rxxxy ...
Rxyyx."
With this information new Users could run a trial analysis without any
necessary knowledge of your hypothesis, but enough information about
your stimulus parameters to run practice analyses on your data.
If you we be happy to provide this information about your data we would
be most grateful. We would then put this in some shared space for all
Users to access. If however you would like to give a more detailed
account of your experiment, then this could contribute towards the
initiative that Tony prompted in the email copied below.
With thanks,
Michael
--
Dr Michael Simpson
Science Liaison Officer
York Neuroimaging Centre
Innovation Way
York
YO10 5DG
Tel: 01904 567614
Web:
http://www.ynic.york.ac.uk
*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_
Copy of mail from Tony Morland date 25th October.
Dear All,
Since the first beamforming course a number of people have been asking
questions like;
What can beamforming really do?
If I do this and get that, what does it mean?
Is beamoforming precise enough to distinguish between activity in brain
area A and B?
If I do X will it result in data that can be analysed with beamforming?
Can virtual electrode traces allow latencies to be measured in brain areas
A and B and do they make sense?
If I filter in this frequency band rather than that band why do I get this
result?
The answers to many of these questions are empirical and although Will et
al have been striving to test, retest and validate the methods, there will
always be questions that can't quite be answered unless more data is
acquired/analysed. It appears necessary therefore to start bench marking
the beamforming methods that are available. The aim is to undertake a
series of empirical studies that will reflect, in broad terms, the types
of paradigm used by investigators. My personal preference would be for a
set of basic experiments to be devised for which results are predictable
and could be compared with previous research. I accept that this is a
conservative aim, but the data sets acquired are likely to be the most
informative about the capabilities of the beamforming analysis tools. The
data acquired would also offer us an excellent resource for training any
researchers wishing to undertake investigations using beamforming.
Before devising new experiments it would be worth reviewing previous
studies undertaken at YNiC to see whether they can be used to bench mark
the beamforming methods. To this end, I would be grateful if you could
provide me with details of data that you think are suitable for bench
marking beamforming. The most helpful form for the information would be
the following:
Project Number and Title
Number of participants and their registered numbers
Paradigm used
Relevant research (the study replicated)
Predicted outcome
What you want to test with beamforming
Ethical concerns/limitations concerning use of the data
I will review the forthcoming information and after that will arrange a
brainstorming session on how best to move things forward.
Thanks
Tony