Deformation algorithms

[AdamY]

The validation work of many algorithms that employed in DIRART is on
going on behalf of my research project.

Optical flow and Demons algorithms are OK to run, but 'Level set
motion method' and 'Methods from ITK' such as B-spline are not
activated, so that I can run it on my imaga data set.

Here are the questions...

1. I would like to test all the algorithms employed in DIRART. Is
there any way I can make all algorithms in DIRART activate to run,
such as 'Original level set motion method' or 'B-spline method'??

2. I used to load file to DIRART via CERR, but suddenly some 'dose
plan' file can be loaded on the CERR to make CERR plan. It gives
error : "cannot generate DVH". Do we need to essentially contain DVH
info into the Dose file when we generate the plan file using TPS??

I look forward to hearing from you.

Kind regards,

Adam Yeo

[Yang, Deshan]

Adam

For your question #2, no, you certainly don't need to have DVH data in
your dose DICOM file.

Is it a CERR problem? To verify it, you should temporarily clear all the
DIRART folders from your MATLAB path (to make sure only CERR MATLAB
files are used) and try to load your plan into CERR (without anything to
do with DIRART).

Here are the questions...

Kind regards,

Adam Yeo
The materials in this message are private and may contain Protected Healthcare Information. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

[AdamY]

Hmm, I just have converted the dicom file to mat file using dicomread
function contained in Matlab imaging process toolbox.
Do we need to keep updating CERR version? (I am using 2010b Matlab).

Any proper answer for the question #1? I really wonder how to run all
DIR algrithms employed in DIRART:)
Thanks for your help in advance.

Adam YEO

> The materials in this message are private and may contain Protected Healthcare Information. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.- Hide quoted text -
>
> - Show quoted text -

[Yang, Deshan]

I will take a look today. The level set algorithm should work. I probably have no time to work out the b-spline algorithm from ITK.

[Yang, Deshan]

I tested the "original level motion algorithm". It works for me.

What are the other algorithms not working for you, besides the b-spline algorithm?

Did you update your DIRART programs from googlecode? There were a few missing functions for level set algorithms. I have added them.

Deshan

-----Original Message-----
From: dir...@googlegroups.com [mailto:dir...@googlegroups.com] On Behalf Of AdamY

Sent: Thursday, June 02, 2011 12:23 AM
To: dirart
Subject: Re: Deformation algorithms

[AdamY]

Yes, I have updated DIRART from googlecode, the 'original level set
motion algorithm' seems to be working :)
The other algorithms not working are the 'original Lucas-Kanade
optical flow method', 'improved Lucas-Kanade optical flow method',
'Hiearchial level set motion method', 'SSD minimization method', and
so on.

I am trying to validate each algorithm using CT data set from
'deformable phantom' with fiducial markers in by comparing the
difference of each fiducials position between 'real measured' and
'computed' 3D array.
It is great to get the 'level set motion' algorithm, but it is not the
best case of not being able to test the 'b-spline algorithm'.

Hope keep in touch with you regarding this DIRART work.

Best wishes,

Adam YEO

[Yang, Deshan]

OK, I have fixed a few problems, and most importantly, fixed the problem about registration algorithm mapping so to allow all broken algorithms to work again, including the mentioned algorithms: 'original Lucas-Kanade optical flow method', 'improved Lucas-Kanade optical flow method', 'Hiearchial level set motion method', 'SSD minimization method', etc.

But sorry, the B-spline algorithm won't work. It was done by someone else and I cannot fix this problem by myself. There are enough implementation of B-spline algorithm available out there, using google to search and you should be able to find them.

Please update your code using subversion from googlecode.