| Re: Registering CT and MRI volumes [message #24054] |
Mon, 05 March 2001 10:01 |
davidf
Messages: 2866
Registered: September 1996
|
Senior Member |
|
|
karthik balasubramaniam (clunis_immensus@MailAndNews.com) writes:
> I remember spotting a thread on registering medical volume data,
> a while back. Unfortunately, I haven't been able to retrieve it at deja.com,
> which I've been using so far(only shows me recent postings).
> Anyways, I need to register CT and MRI volumes, and was hoping someone could
> point me to IDL(or even C/C++) code that does volume registration.
> I do have some literature on the topic, but thought it worthwhile sounding
> out
> people before reinventing the wheel.
> Any links, references, code would be immensely appreciated.
I have in front of me a pre-print of an article entitled
"Volumetric image registration by template matching", written
by Ardi Goshtasby's group at Wright State University. I saw
this algorithm in action several weeks ago, and I must say
I was extremely impressed with the speed and accuracy they
have obtained.
The basic idea is to use a handful of templates from
"high information content" areas of the volumes to
do the matching, rather than matching the entire
volumes. In this case, "high information content" means
edges, etc. The algorithm can automatically select
these templates and perform the matching. I saw it
work well when the regions were just slightly mismatched
and when they were significantly mismatched. I was
impressed.
I understand the IDL algorithm will be made available for
research purposes. You can write to Ardi Goshtasby at
agoshtas@cs.wright.edu for additional information. Please
give him my compliments.
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Registering CT and MRI volumes [message #24063 is a reply to message #24054] |
Sun, 04 March 2001 12:43  |
Randall Frank
Messages: 17
Registered: October 1999
|
Junior Member |
|
|
Karthik,
One "standard" algorithm is Roger Wood's AIR:
http://bishopw.loni.ucla.edu/AIR3/index.html
Many of the more popular (particularly functional)
imaging apps use variants of this approach. I have not
seen it coded in .PRO yet (it actually could be done fairly
efficiently), but the technique is very straight-forward
and reasonably robust for small corrections. The functional
imaging group I work with has had good results with it.
I have been thinking about a writing a DLM for it, but I was
waiting until the next release which I hope will allow
real-time use of the algorithm. My application is real-time
fMRI. We use IDL (augmented with a lot of DLMs):
Smyser C, Grabowski RJ, Frank RJ, Haller JW, Blinger L,
"Real-time multiple linear regression for fMRI supported by
time-aware acquisition and processing", Magnetic Resonance
in Medicine, Volume 45, Issue 2, 2001, pp. 289-298.
Not exactly what you were asking for, I hope you find
some of it useful.
BTW: on your KB_ISOSURFACE, might I suggest adding a DLM
to compute an "optimal" complete seed set for a given
dataset? The combination of that and your KB_ISOSURFACE
would yield a complete, high-performance replacement for
the existing IDL DLM. I use a variant of this myself for
N-D isocontouring. A good reference might be:
Bajaj CL, Pascucci V, Schikore DR,"Fast Isocontouring For
Improved Interactivity", 1996 Volume Visualization Symposium,
ISBN 0-89791-741-3, pp.39-46
Apologies for my long-winded response. Those of you who
know me, know I can not help it...
karthik balasubramaniam wrote:
>
> reply-to: clunis_immensus@hotmail.com
>
> I remember spotting a thread on registering medical volume data,
> a while back. Unfortunately, I haven't been able to retrieve it at deja.com,
> which I've been using so far(only shows me recent postings).
> Anyways, I need to register CT and MRI volumes, and was hoping someone could
> point me to IDL(or even C/C++) code that does volume registration.
> I do have some literature on the topic, but thought it worthwhile sounding
> out
> people before reinventing the wheel.
> Any links, references, code would be immensely appreciated.
>
> Also, could someone recommend me a web-based newsreader where I could access
> old postings of this newsgroup?
>
> many thanks,
>
> Karthik Balasubramaniam
> CAS Technologies, Chennai
>
> ------------------------------------------------------------
> Get your FREE web-based e-mail and newsgroup access at:
> http://MailAndNews.com
>
> Create a new mailbox, or access your existing IMAP4 or
> POP3 mailbox from anywhere with just a web browser.
> ------------------------------------------------------------
--
rjf.
Randy Frank | ASCI Visualization
Lawrence Livermore National Laboratory | rjfrank@llnl.gov
B451 Room 2039 L-561 | Voice: (925) 423-9399
Livermore, CA 94550 | Fax: (925) 423-8704
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
