RTOG import crashes with images that have different x and y resolution

[Ed Bender]

Hi,

I am trying to figure out why the RTOG import is crashing when I try
to import the ICBM brain image (it's an MRI image of a brain). The
unfortunate thing about this image is that the resolution is 181x217
with 181 slices. I don't know why the people who made this image
decided to use such a ridiculous data size but it's what I'm stuck
with.

If I don't modify the code I get a subscripted assignment dimension
mismatch in the "Build the 3-D scan matrix" block of the CERR/
Importing/importRTOGDir.m code, at the line scan3M(:,:,i) = CTM;

I've tried messing with the code and I can only get it to sort-of
work. If I remove the fliplr at the line CTM = ctin(optS, fileStr,
fliplr(CTSizeV));, it doesn't crash but I get a scrambled image.

If I remove the line CTM = CTM'; It will work but the image comes in
transposed

So what I can get is either a scrambled or transposed image but not
the correct image. Trying to manually transpose the image after
importing doesn't work because of the non-equal x and y dimensions.
Any ideas? It seems like it's something silly like a fliplr or
transpose, or both, but nothing I've tried has worked.

Thanks,
-Ed

[Apte, Aditya]

Hi Ed,

Thanks for reporting this. CERR is designed to handle unequal resolution, unless there is a bug. I do not recall whether we have tested RTOG import for MRI scans before. Is it possible to share the data with us? If so, can you please upload the anonymized data to our server or some location? We will look at it and let you know.

Best regards,
Aditya

--

You received this message because you are subscribed to the Google Groups "CERR FORUM" group.
To post to this group, send email to cerr-...@googlegroups.com.
To unsubscribe from this group, send email to cerr-forum+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/cerr-forum?hl=en.

 

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.

[divya khullar]

Ed,
Check if all your images are equal size; by that I mean X & Y dimension. There might be two data set that you are trying to import. If you do provide data, Aditya can give you a better diagnosis.
- Divya 

[Ed Bender]

To be more specific, the voxel size is uniform (0.9mm in all
dimensions) but there are different numbers of pixels along the x and
y dimensions. Each slice has the same number of pixels if that's what
you're wondering.

How do I upload to your server? My email is bend...@yahoo.com if you
need to tell me privately.

[David Turnbull]

Hi Ed,

Did you get your problem solved? That code is a bit wacky with the
constant dimension flips. I can see a fix but have no way of testing
it.

Cheers,
Dave

On Dec 12 2009, 7:49 pm, Ed Bender <bende...@yahoo.com> wrote:
> To be more specific, the voxel size is uniform (0.9mm in all
> dimensions) but there are different numbers of pixels along the x and
> y dimensions.  Each slice has the same number of pixels if that's what
> you're wondering.
>

> How do I upload to your server?  My email is bende...@yahoo.com if you

[Ed Bender]

No, I haven't gotten it fixed. I ended up taking a back door approach
to what I was working on. If you have a fix I can try it in no time -
just send me whatever edited code you want and I'll give it a whirl.

Thanks,
-Ed