Dear Shoichi:
I am still confused about your model, but here is my guess.
Input effects are on regions (DCM.c matrix). Intrinsic connections are
between regions (DCM.a), and modulatory effects are on the intrinsic
connections (DCM.b)
So from what you have said, it appears that you have have 1 input condition,
that of fearful faces and 0 modulatory effects.
I assume that in the matrices o = 1 and x = 0; by the way it would be easier
to read if you send the matrices as 1's and 0's.
DCM.c = [1 1 1 1 0]'; (inputs to all regions EXCEPT the PFC)
In general one might consider just 1 or perhaps 2 input areas. If inputs go
to all, or nearly all areas, it turns the DCM analysis into somewhat of a
conventional SPM analysis in the sense that activity in SPM's are analogous
to stimulus input to all regions (see Friston et al., Neuroimage, vol 19,
page 1275 (bottom first - top second column), 2006.
DCM.a = [1 1 1 1 0
1 1 1 1 0
1 1 1 1 1
1 1 1 1 1
0 0 0 0 1]
DCM.b = [0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0]
If I have read the matrices correctly the other problem is that you have
inputs into all areas except PFC, none of the areas connect to the PFC, but
the PFC connects to R and L amygdala. In that case, PFC is receiving no
inputs either directly or indirectly. Thus it can produce no output.
Darren.
-----Original Message-----
From: Shoichi Ugai [mailto:[log in to unmask]]
Sent: Wednesday, November 15, 2006 6:45 PM
To: Darren G
Cc: Shoichi Ugai
Subject: Re: DCM model
Dear Darren
Thank you for your reply.
I am sorry but there was a mistake in my previous description.
> the effect of fearful face are:
> R.FFA o x x x x
> L.FFA o x x x x
> R.AMY o x x x x
> L.AMY o x x x x
> PFC x x x x x
is not correct.
Previous my connetion was
> R.FFA o x x x x
> L.FFA o x x x x
> R.AMY o x x x x x
> L.AMY o x x x x x
> PFC x x
However, we got the results as stated previously(wrong DCM.A).
We hypothesized that the effect of fearful face directly affect the
R.FFA, L.FFA, R.AMY, and L.AMY.
Is this not enough?
If we add the connection as follows, we also get strange results.
R.FFA o x x x x
L.FFA o x x x x
R.AMY o x x x x o
L.AMY o x x x x o
PFC x x
Thank you again,
Shoichi
|