View Full Version : Summary of CT scans to 3D models

Julie Matthews
11-25-1997, 01:16 AM
Dear Biomecher's
Thanks to everyone who took the time to reply to my question:

For the purpose of improving our patient care, we investigating methods of
making 3D models of long bones from our patients CT scans.

I have looked at the SurfDriver software, but find the method of
highlighting the areas of interest, to be slow and error prone
Hence I was wondering if anyone could suggest or recommend:

1)A program that can 'atomatically' process CT scans.

2) An alternative to the surfdriver.

The informative and helpful replys are listed below:
I am sure the software you are looking for is called Voxar. I am
testing it at the moment and it looks promising. You need to get in touch
with Ms Alison Sinclair at 44-131-4724792 (Edinburgh, Scotland) or
visit www.voxar.com.

good luck.
Yours sincerely,

Dr Rami J Abboud, BEng, MSc, PhD
You may want to try Fraunhofer. It is a non-for-profit organisation that
lots of knowledge in this area.

The contact that I have is Paul Wehner at pwehner @ fraunhofer.bu.edu .

He is located at Boston University. I do know that Fraunhofer has may labs
around the world. Paul may beable to direct you to one closer to you.


Patti Turnbull
Email patti.turnbull@reebok.com
You may wish to contact Dr. Joyce Keyak (keyak@ortho.ucicom.uci.edu),
who has had an algorithm in the literature for automated finite
element model generation from quantitative CT scans since about
1990. How useful it is might depend on exactly what you plan on
doing with these models.
Cliff Les, DVM, PhD
Email cliff@equus.ucdavis.edu
There is a program called Med-link that runs in Ideas that takes CT
data and creates 3-d models from it. You also do stress, thermo, etc..
analyses also. http://www.primenet.com/~dcr_inc/
Brian G.
Email Engeniero@aol.com
We are actively engaged in this issue. I agree
with you that you must calibrate your segmentation
and isosurface building routines in order to
produce accurate models via STL. We are actively
engaged in this work. Did you attend the recent
meeting in Erlangen:


I would be happy to discuss this further.


|David Dean, Ph.D.
E-mail: dxd35@po.cwru.edu |
An alternative to surfdriver is http://www.materialise.com
I can strongly recommend you the software of Materialise, a company here in
Leuven (Belgium) that is specialised in software for the processing of
medical images (CT, MRI). They have recently developped a software
programme called MedCAD, that presents a direct link between medical images
and CAD. With this programme you can easily convert anatomical structures
(like e.g. long bones) to a surface description of the outline of the
anatomical structure. You can read in medical images that comes from
almost any kind of scanner (Siemens, General Electric...). MedCAD then
works as follows : first you have to apply a certain grey value threshold
to segment your images. MedCAD then automatically calculates the contours
that outline your segmentation object for each images slice. With these
contour data a spline surface is automatically calculated. MedCAD exports
this surface data in an IGES format, so you can import the surface in a CAD
programme. I have experimented with the programme myself and I must say
that the results are really great. You can get a high accuracy to
approximate even very thiny irregularities of the anatomical structure. A
colleague of mine has used MedCAD for the construction of a 3D surface
model of a CT-scanned tibia and she was able to reconstruct the geometry
very accurately with only one spline surface. I use MedCAD for the
reconstruction of 3D models of the upper and lower jaw from CT. For these
complex geometries I end up with a number of splines which I can unite in
a CAD programme.
If you want additional information you can visit Materialise's website at
www.materialise.be. I don't know whether you can get a demo version for
evaluation purposes nor what the prices are for the MedCAD software.

With friendly greatings,

Hans Van Oosterwyck
e-mail: hans.vanoosterwyck@mech.kuleuven.ac.be
a few weeks ago, a summary concerning 3D analysis software has been
posted. Perhaps this might help you.

I am working on the modelling of lower human extremity, so I have to
build the model (from the visible human data set), including bones,
muscles and ligaments. I have ordered a licence for the SURFDriver
software, because the demo version seemed to me quit good. Perhaps you
can tell me more about the experience you have made with this software.
If you get further replies, perhaps you can send them to me, too.


email pohl@pollux.medizin.uni-ulm.de

Thanks to everyone.

Julie Matthews. M.Sc Bioengineer
The Institute for Applied Biotechnology
Medicinaregatan 8B
413 46 Göteborg, Sweden
Phone 46 (0)31-416337 or 411241
Fax 46 (0)31-414560
email Bioteknologen@swipnet.se

To unsubscribe send UNSUBSCRIBE BIOMCH-L to LISTSERV@nic.surfnet.nl
For more information: http://www.kin.ucalgary.ca/isb/biomch-l.html