This is unrelated to your question, but you might find useful that, if you
increase your TE from 20 ms to 40 - 50 ms on your next experiments, your
BOLD contrast signal will increase significantly (I'm assuming it's a 1.5
T magnet). There are several studies, showing this TE dependence (I don't
recall the references, but if you are interested I can dig them up), and it
made a big difference for us. I hope this is helpful.
Luis,
------------------
Luis Hernandez, Ph.D.
Research Fellow
Dept. of Neurosurgery
Wake Forest University School of Medicine
Medical Center Blvd.
Winston Salem, NC 27157
Phone: (336) 713 8068
Fax: (336) 713 8588
-----Original Message-----
From: John Ashburner [SMTP:[log in to unmask]]
Sent: Thursday, October 01, 1998 5:14 AM
To: [log in to unmask]; [log in to unmask]
Subject: Re: realignment problem
I would check the ORIGIN's to see if they are written correctly
by the conversion process. The realignment uses the relative
positions of the ORIGIN to obtain starting estimates for the
registration. If the starting estimates don't suggest enough
overlap of the images, then the realignment starts producing
rank deficient warnings.
If it's not the ORIGIN field, then it could be the VOX field,
or possibly some ".mat" files that contain wrong information.
Good luck,
-John
> I am analyzing fMRI data using spm.
> Our data were required by GE signa with ANMR system.
> Parameters were TR=4s, TE=20ms and flip angle=90, and
> Acquired images were 131 data points and slices were
> 15 coronal slices.
> We transformatted these images to ANALYZE format
> using self-made software.
>
> We tried to realign these prepared images using spm.
> But, the following message was presented:
> "Warning: Rank deficient, rank=0 tol=?.0000e+00"
> After this message was presented, "r*.img" files
> were not made.
>
> How can I solve this problem?
> Please give us useful suggestion.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|