View Full Version : FEM of bone

Daniel A. Michaeli
02-07-1994, 10:37 PM
Recently, I sent a message requesting any information regarding FEM modelling
of bone from CT. Specifically, I was interested in the assignment of material
properties and an edge detection algorithm. I am sending the results
since I received many replies from those who were interested in the
result of my query.

************************************************** ***************************

You might like to contact:

Laurence Marks
Desktop Engineering Ltd
Evenlode Court
Long Hanborough
Tel. 011-44-993-883555

I know he is developing just such a technique, but you would have to pay!

Robert Scott

************************************************** *************************

I don't have any really quick or automated methods, but I have created one
2-D mesh using NIH Image (which has some helpful thresholding tools) and
I-deas (SDRC) FEM software. I would however be interested in any postive
responses you might get. Please, send them my way or post them back to
the biomech-l listserver!!!


Ken Fischer
Stanford University

************************************************** *************************

Check out some software from the Cornell University Hospital for Special
Surgery which was designed to do just that. Academics can purchase for
about $3000. to run on UNIX workstations. File format for input to this
package is an issue, but that will always be an issue with CT files.
Contact the software group at Cornell via:

Don Bartel
Dean Taylor
219 Upson Hall,
Cornell University
Ithaca, NY 14853
(607) 255-0990


Good luck.

Don Anderson, Ph.D. -- Director, ASRI Biomechanics Research Laboratory
Pittsburgh, PA

************************************************** ***********************

Thanks for your acknowledgement. Let me also mention that SDRC has an
office in Littleton MA. Steve Blanchette is the District Manager -- phone
508-486-9883; fax -0518.


Tom Sigafoos

************************************************** ***************************
1) Joyce Keyak at UCSF (jkeyak@garnet.berkeley.edu) has developed an
automated FEM generator from serial CT scans. She and Harry Skinner
(among others) have published extensively on this method since about
1989, and so far they've done human femora, vertebrae, and shoulders.
See ORS proceedings for the last few years for a start.
2) Joyce and I have got a paper in the mill at JOR (accepted last fall,
God only knows when it'll be published) relating K2HPO4 equivalent
density from QCT with elastic modulus, yield/failure strength, stress,
density, the works, for bone material from the equine metacarpus.
For human bone you might want to start with Snyder and Schneider,
JOR 9(3) :422-31, 1991
Good luck
cliff les
************************************************** ***************************
Someone in my lab gave me a copy of your email about developing finite
element models of bone quickly from CT. It so happens that I have
been working on that for a while. One paper you might look at
is Keyak, Meagher, Skinner and Mote: Automated three-dimensional
finite element modelling of bone: A new method. J Biomed Eng
12(5):389-397, 1990. I think that paper should answer most of
your questions.

Edge detection: The method that I have been using is described by
Seitz and Ruegsegger (referenced in the above paper). I don't
know if the contours can be imported into conventional FEM programs.
You could presumably format the contours however you need them.
The problem is whether the FEM program can deal with the bone
geometry, and this would depend on which bone you want to model.

Material properties: As you probably know, it is fairly straight-
forward to convert CT attenuation to physical density, and physical
density to elastic modulus. (Of course, there are numerous
assumptions made in the process, and there is considerable uncertainty
in predicting elastic modulus.) The time consuming part is to
get the CT densities and predicted moduli mapped to the elements.

With the program I developed, we can generate a finite element
model of a bone in well under a day, after we have the CT data
uploaded to the computer. The latest upgrades have it running
in a very user friendly way. Two clicks of the mouse
will give you a contour. If you find yourself in San Francisco,
I would be happy to help you generate a model. However, I would
not feel comfortable sending you the program. I am sure you
understand. Whether the approach that my program uses is for you,
would depend on your particular application (see the above paper,
and you will see what I mean).

You may be aware of the program developed at Cornell. It is
supposed to be quite slick, from what I have heard. This program
also is supposed to be able to do what you are asking. They
use a somewhat different approach, so it would take a bit
longer to generate the model. There are tradeoffs. The method
I use is better at some things; their method is better at
others. See: Keyak et al: Validation of an automated method
of three-dimensional finite element modelling of bone. J Biomed
Eng 15:505-509, 1993. This paper gives the pros and cons of
the two methods.

I hope this information is helpful to you. Let me know if
I can be of help. I would be interested to know what you decide
to do.

Joyce Keyak
************************************************** ************************