[Brains-users] RE: Thalamic MD measures and greeting

[Ronald Pierson]

Hi, this is Ron Pierson.  I trained the ANN on that structure, and it really
was pretty poor with ICC of ~0.5.  The ANN method created as much
variability in the data as there actually is in the sample.  There just was
too much variability in the probability maps, and too little contrast at the
MD borders, so the ANN didn't have good information to start with.  We just
would not believe any findings that were based on that bad data.  There are
some other ANN structures that are also pretty bad, but they serve as a good
starting place for editing, which we always like to do anyway.  This is not
the case for the MD.  So, we decided not to release it.  Really, I think the
best at the moment would be a manual method, though it may be rather
labor-intensive to get the experience needed to do it.

I think we look forward to a couple of new methods in the next year or so
that may do a much better job on some of these structures.  I'm sure we will
announce on this list when they become available.

See you in Florence?

Ron


-----Original Message-----
From: Benedicto Crespo Facorro [mailto:bcfacorro at humv.es] 
Sent: Wednesday, February 22, 2006 10:10 AM
To: hans-johnson at uiowa.edu
Cc: RobertoRoiz
Subject: RE: Thalamic MD measures and greeting

Dear Hans, 

My name is Benedicto Crespo-Facorro and I was a former postdoc at the IPL
IC. 
We are exploring thalamic volumes. We have obtained total thalamic volume.
We are now trying to know whether it is possible to get automated measure
(neuro net) of the Thalamic MD, as Ruth published in NI 2002.
Could you help out us to find where is the option in BRAINS2 menu?

Thanks for your time and interest,
bn





Ruth,
Thanks and good luck with your research too!!
bn

-----Mensaje original-----
De: Spinks, Ruth A [mailto:ruth-spinks at uiowa.edu] 
Enviado el: miércoles, 22 de febrero de 2006 16:45
Para: Benedicto Crespo Facorro
Asunto: RE: Thalamic MD measures and greeting


Bene,

Good to hear from you and I'm glad to hear your imaging research is doing so
well. I would suggest you ask Hans Johnson whether there is an MD neural
net.  I am no longer working in the CRC so I haven't been doing imaging for
the last 3 years. I took over Remi Cadoret's adoption study and I work with
Steve Arndt doing research on the cognitive effects of substance abuse and
dependence.

Good luck.
Ruth

-----Original Message-----
From: Benedicto Crespo Facorro [mailto:bcfacorro at humv.es] 
Sent: Wednesday, February 22, 2006 9:35 AM
To: Spinks, Ruth A
Cc: RobertoRoiz
Subject: Thalamic MD measures and greeting

Hi Ruth,
Hope you are doing well and your research successful.
As you may know we have been collected MRI in the last three years (FES).
Just recently we started to analyze data. We found a decrease in total
thalamic volume in FES (79) when comparing to healthy volunteers (45) . 
I read your paper in NI regarding the method. In that paper you described MD
measures. Is this option (MD neuronet) available in BRAINS2 ?? We can not
find it in the menu. Thanks a lot and warm hugs,

Bene

Benedicto Crespo-Facorro MD,
Schizophrenia Research Program
Dep. of  Psychiatry. Pabellón 20.
HU Marqués de Valdecilla
39008 Santander
Spain