PDA

View Full Version : Summary of Responses: Software for 3D reconstruction from MRIscans



Bhushan Borotikar
12-15-2005, 08:07 AM
Dear Biomech-L Subscribers,

Thank you everyone who responded to my query on software for 3D reconstruction from MRI scans. Responses mainly suggested two software names “MIMICS” and “AMIRA”. I also searched another software called “Analyze” which might do all the tasks I need. Currently I am in the process of evaluating Amira and Analyze. I also received some good suggestions and tips about handling the MRI data. My original query is pasted below followed by responses.

Regards,

Bhushan S. Borotikar
Department of Biomedical Engineering
Lerner Research Institute (ND1-08)
Cleveland Clinic Foundation
9500 Euclid Avenue, Cleveland, OH 44195
Phone: (216) 445-5607


Original Query:

Greetings Biomech-L subscribers,
I am a graduate student in Lerner Research Institute. I am currently working on developing a knee joint model for my project. The articular cartilage surface points and ligament insertion points are digitized from MRI scans.
The ligament insertion points are the most difficult ones to get from these scans. I am looking for software that would take MRI scans from all the three planes (sagittal, axial and coronal) and make a 3D volume.
Software should be able to rescan this volume at any oblique plane. This way I can exactly locate the ligament insertion points and can accurately digitize them (get the x, y and z coordinates).
I am wondering if anybody knows or using this kind of software package. I know a few of them (like 3D Doctor), but they take stack of MRI scans in single plane to make the volume.
I will post the summary of responses in due time. Thank you very much in advance for any help with this request.

Regards,

Bhushan S. Borotikar
Dept of Biomedical Engineering
LRI, Cleveland Clinic Foundation

-------------------------------------------------------------------------------
Response # 1
-------------------------------------------------------------------------------

Bhushan,

I believe that Amira can do this. I have used Amira in the past to recreate the meniscus geometry using scans from multiple planes.

Steve Fening, PhD
The Cleveland Clinic Foundation

--------------------------------------------------------------------------------
Response # 2
--------------------------------------------------------------------------------
Bhushan-
I would look at the University of Virginia (Jason Kerrigan-PhD student/ Jeff Crandall -Mechanical Engineering Professor) and their knee ligament human subject impact testing and modelling. I think they have used a knee model that incorporates origin and insertion locations.
Hope this helps,
Adam Bartsch
Sealimited.com
---------------------------------------------------------------------------------
Response # 3
---------------------------------------------------------------------------------

Bhushan
I have done a little of this sort of work, with CT rather than MRI data. I have used AMIRA and most recently MIMICS (I am still finding my way around in it). I am not an overly experienced user, so these comments may be influenced by my limited experience:
AMIRA seems quite powerful. It is able to create oblique planes. Command line and scripting are available if you are doing repetitive tasks. I have not explored the full scope of AMIRA yet, so there may be things that it does that I am not aware of yet.
MIMICS also seems powerful. I like its graphical interface better than AMIRA, however, I am not sure that there are scripting capabilities. I assume that it can create oblique planes, I have not tried this yet. MIMICS seems to be more integrated and has more options for 3D output for downstream analysis. I am not aware of price.
Good luck
Andrew Sims
Graduate School of Biomedical Engineering
University of New South Wales
---------------------------------------------------------------------------------
Response # 4
---------------------------------------------------------------------------------
Bhushan,
I used Materialise MIMICS software to make a three-dimensional reconstruction
of a spinal unit from CT scans. Take a look at their website
http://www.materialise.com/mimics/main_ENG.html
or give them a call
at 1-734-662-5057. Hope this helps.
David Feldstein
BS/MS in Biomedical Engineering
Drexel University

---------------------------------------------------------------------------------
Response # 5
---------------------------------------------------------------------------------
Mimics is the one we use in our lab.

Nicole Wilson
Illinois Institute of technology

---------------------------------------------------------------------------------
Response # 6
---------------------------------------------------------------------------------
amira (http://www.amiravis.com/) should do it.
regards,
chris
Dipl.Ing. Christian König
Center for Musculoskeletal Surgery
- Research Laboratory -
Charite - University Medicine Berlin
Humboldt-University and Free University
Augustenburger Platz 1, D-13 353 Berlin
tel: +49.30.450.552097 (Office)
tel: +49.30.450.552012 (Dept.Sec'y)
fax: +49.30.450.559969
email: christian.koenig@charite.de
www: http://www.charite.de/cmsc

---------------------------------------------------------------------------------
Response # 7
---------------------------------------------------------------------------------
Bhushan -

I am also interested in this, for dynamic reconstruction of lung tissue for flow modeling. I look forward to seeing the responses to your request. Unfortunately I am only learning about this field myself, and I do not have any suggestions for you. I am glad you asked the question, because you saved me the effort!

Are there any other people in your research institute working on this sort of thing? I have not heard of LRI before...

Kind Regards,
Karl Wittnebel
Washington, DC


---------------------------------------------------------------------------------
Response # 8
---------------------------------------------------------------------------------
Dear Mr. Borotikar,
Perhaps this software can have the functions for your research:
http://www.rapidform.com/17
Tom Charron
tomc@rapidform.com
Regional Manager - Americas
RapidForm 3D Scanning Software
---------------------------------------------------------------------------------
Response # 9
---------------------------------------------------------------------------------

Hi,
I just used a little "MIMICS" this is the name of sofware. I know it is difficult task to reconstruct from MRI scans.
Good luck.
Hossein Mokhtarzadeh Salmasi.
School of Mechnaical Engineering,
Sharif University of Technology,Tehran,Iran
http://sina.sharif.edu/~biomech
---------------------------------------------------------------------------------
Response # 10
---------------------------------------------------------------------------------
We used MIMICS from materialise to create 3D from CT, its quite good and works with MRI as well.
Kate
University of Aberdeen
Aberdeen, Scotland
---------------------------------------------------------------------------------
Response # 11
---------------------------------------------------------------------------------
Hi Bhushan,
There are two software packages that I know of: ScanIP by Simpleware and Mimics by Materialise.
Both are available here at BME.

Azita Tajaddini, D. Eng.
Cleveland Clinic Foundation

---------------------------------------------------------------------------------
Response # 12
---------------------------------------------------------------------------------

Hello,
This is not a direct answer to your question, just a comment that might be of use. Keep in mind that some ligaments wrap around bone (some band of the MCL and the LCL). If you use straight lines to represent the ligaments and if you are not careful defining your insertion points, you can end up with an anatomically incorrect ligament running through bone.
Regards,
Eihab Abdel-Rahman, PhD
Virginia Tech
---------------------------------------------------------------------------------
Response # 13
---------------------------------------------------------------------------------
Hello Bhushan,
My thesis work involved the use of MRI data to recreate the knee.
I encountered the same problems that you seem to be finding with respect to articular cartilage and soft tissue demarcation from bone and other soft tissue. Mimics by Materialise (www.materialise.com) was the software package that I utilized at the time. I believe that the package has become more accurate
with respect to the aforementioned problem. They did allow for trial versions prior to purchasing and you may want to look into using Their software package. You should also be aware that the type of MRI method utilized is important for viewing soft tissues and the demarcations between them. I found that the 3D Fat-suppressed SPGR imaging technique yielded the best pre-processing
images.
Hope this helps.
Best regards,
Naphysah Duncan MSc.
Project Engineer
Cooper Union
51 Astor Place
New York, NY 10003
Phone: (212) 353 - 4042
Lab: (212) 353 - 4326
Fax: (212) 353 - 4038
---------------------------------------------------------------------------------
Response # 14
---------------------------------------------------------------------------------
Hi Bhushan,
I faced the same problem when I built my finite element model of the knee joint. I found using Matlab and its imaging processing toolbox helpful, the only downside is some programming. Basically, the way I approached this problem is first generating a 3D volume from the single plane MR images using "cat" function,
you might also want to resize the image to make the pixel size the same in three directions (function "imresize").
Secondly, you can retrieve a 2D matrix from the 3D volume matrix in what ever way you want, whether it is sagittal plane or coronal/axial planes using function like "squeeze".
Hope it helps..
Jiang Yao
Department of Mechanical Engineering
Univeristy of Rochester
---------------------------------------------------------------------------------
Response # 15
---------------------------------------------------------------------------------
If you happen to get any good responses from this could you forward them to me? My lab is in the process of setting up an MRI machine and we'd like to be able to get 3D reconstructions of the lower limb to get data like muscle attachment points, limb geometry, and other data for kinematic anaylsis.
Thanks
Brian K. Hillen
Brian.Hillen@asu.edu



---------------------------------
Yahoo! Shopping
Find Great Deals on Holiday Gifts at Yahoo! Shopping