Dear Mark,
Thanks for your detailed reply. While i was waiting for reply to my question i tried registering my my slice gap image with the other image i.e. image_2 (contiguous) slices it worked very well, the registration was good, my image_1 with slice gap has better contrast when conpared to Image_2 , so i used my image_1 as reference as i metioned to Steve, i didn't see any problems in my registered image. However as i am new to image processing ( i don't know much) i would like to follow your suggestion and see whether are there any differences between the two ways. So in this regard i would like to clarify the following with you
Step 1 - create an image with alternating slices of
your image and an extra copy of these slices,
where each slice is 4mm thick
e.g. slice1 = orig slice1, slice2 = orig slice1,
slice3 = orig slice2, slice4 = orig slice2,
slice5 = orig slice3, slice6 = orig slice3, etc...
I understand this.
Step 2 - create a weighting image which is full of ones
in the odd slices and full of zeros in the even
slices (NB: all voxels in the odd slices should be
set to 1 - not just the brain voxels)
I understand this.
Step 3- do the registration between your images using
this weighting volume appropriately
I am not able to understand this step because i hearing "weghting image" idea for this first time here, so can you explain to me on this a little more like after creating this weighting image of ones and zeros, should i have to make this weighting image same size as that of step 1, do it need to multiply this with my image_2 (contiguous slice image).
It might be simpler for the registration code in this case to treat your created image above as the reference image, and hence the weighting image as -refweight.
Note that it is important that you do not put zeros in the even slices for the ref image itself, as the smoothing
would be affected by the zeros in a bad way.
I am sorry that i am not able to understand this step also because you have mentioned not to put zeros in even slices for ref image but in the previous step you have suggested to me to use weighting image as reference image, may this will be clear if i understand what it means by "weighting image",
I am sorry for bothering you,
Thanks,
Regards
venkateswaran
________________________________
From: FSL - FMRIB's Software Library on behalf of Mark Jenkinson
Sent: Thu 3/19/2009 9:49 AM
To: [log in to unmask]
Subject: Re: [FSL] FLIRT on images with gaps between slices
Hi,
If you have huge slice gaps, like you seem to have,
then the simple prescription that we normally use,
and that Steve described, may not work. The
reason is that there is a big difference between what
an image scanned with an 8mm slice thickness
would look like compared to your image. Particularly
in slices containing only a small amount of brain
(e.g. near the top of the brain for axial slices).
Doing the registration where you assume your
image has 8mm thick slices will probably bias the
results (e.g. shifted down).
In order to create a less biased result I would
suggest the following:
- create an image with alternating slices of
your image and an extra copy of these slices,
where each slice is 4mm thick
e.g. slice1 = orig slice1, slice2 = orig slice1,
slice3 = orig slice2, slice4 = orig slice2,
slice5 = orig slice3, slice6 = orig slice3, etc...
- create a weighting image which is full of ones
in the odd slices and full of zeros in the even
slices (NB: all voxels in the odd slices should be
set to 1 - not just the brain voxels)
- do the registration between your images using
this weighting volume appropriately
It probably doesn't matter much which image
you choose as the reference if the intensity contrast
in each is good. It might be simpler for the registration
code in this case to treat your created image above
as the reference image, and hence the weighting image
as -refweight.
Note that it is important that you do not put zeros in
the even slices for the ref image itself, as the smoothing
would be affected by the zeros in a bad way.
This will hopefully work OK, although it isn't something
that we've really tested, and I'm not sure that the
end quality will be brilliant given the large slice thickness.
However, try it and see how you get on.
All the best,
Mark
Rajagopalan, Venkateswaran wrote:
> Thanks Steve for your kind clarification, I will follow your suggestion of choosing a high quality reference image and do my registration.
>
> Thanks a lot.
>
> venkateswaran
>
> ________________________________
>
> From: FSL - FMRIB's Software Library on behalf of Steve Smith
> Sent: Thu 3/19/2009 7:51 AM
> To: [log in to unmask]
> Subject: Re: [FSL] FLIRT on images with gaps between slices
>
>
>
> Hi - when you have slice gaps this should simply reflected in the
> voxel dimension in the header, which should be the summation of the
> slice size plus the gap - i.e., should be the distance from the start
> of one slice to the start of the next. You should not insert gappy-
> data-slices into the actual image.
>
> A separate point is that in general you should use the higher-quality
> image as the reference image, and if necessary then invert the
> resulting transformation before applying that.
>
> Cheers.
>
>
>
> On 18 Mar 2009, at 16:15, Rajagopalan, Venkateswaran wrote:
>
>
>> Dear FSL users,
>>
>> I want to register two images for instance my Reference image
>> is :Image_1 and input image is Image_2 both acquired with same in
>> plane resolution of 0.98*0.98*4mm but in the case of reference image
>> i.e.Image_1 the FOV was chosen to cover only a part of head,so i
>> have only few slices with slice gaps of 4mm thickness introduced in
>> this acquisition, whereas Image_2 on the other hand was acquired
>> contiguously without any slice gap covering the entire head. I want
>> to register my Image_2 to Image_1 space, so my question is can i
>> directly feed them into FLIRT and FLIRT will take care of this slice
>> gap problem during registration or for instance do i need to insert
>> zero padded images for the missing slices in my Image_1 and do the
>> registration.
>>
>> Thanks in advance,
>>
>> Regards
>> venkateswaran
>>
>> ===================================
>> P Please consider the environment before printing this e-mail
>>
>>
>>
>> Cleveland Clinic is ranked one of the top hospitals
>> in America by U.S. News & World Report (2008).
>> Visit us online at http://www.clevelandclinic.org <http://www.clevelandclinic.org/> <http://www.clevelandclinic.org/> for
>> a complete listing of our services, staff and
>> locations.
>>
>>
>> Confidentiality Note: This message is intended for use
>> only by the individual or entity to which it is addressed
>> and may contain information that is privileged,
>> confidential, and exempt from disclosure under applicable
>> law. If the reader of this message is not the intended
>> recipient or the employee or agent responsible for
>> delivering the message to the intended recipient, you are
>> hereby notified that any dissemination, distribution or
>> copying of this communication is strictly prohibited. If
>> you have received this communication in error, please
>> contact the sender immediately and destroy the material in
>> its entirety, whether electronic or hard copy. Thank you.
>>
>>
>
>
> ---------------------------------------------------------------------------
> Stephen M. Smith, Professor of Biomedical Engineering
> Associate Director, Oxford University FMRIB Centre
>
> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
> +44 (0) 1865 222726 (fax 222717)
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
> ---------------------------------------------------------------------------
>
>
>
>
> ===================================
>
> P Please consider the environment before printing this e-mail
>
> Cleveland Clinic is ranked one of the top hospitals
> in America by U.S. News & World Report (2008).
> Visit us online at http://www.clevelandclinic.org <http://www.clevelandclinic.org/> for
> a complete listing of our services, staff and
> locations.
>
>
> Confidentiality Note: This message is intended for use
> only by the individual or entity to which it is addressed
> and may contain information that is privileged,
> confidential, and exempt from disclosure under applicable
> law. If the reader of this message is not the intended
> recipient or the employee or agent responsible for
> delivering the message to the intended recipient, you are
> hereby notified that any dissemination, distribution or
> copying of this communication is strictly prohibited. If
> you have received this communication in error, please
> contact the sender immediately and destroy the material in
> its entirety, whether electronic or hard copy. Thank you.
>
>
===================================
P Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals
in America by U.S. News & World Report (2008).
Visit us online at http://www.clevelandclinic.org for
a complete listing of our services, staff and
locations.
Confidentiality Note: This message is intended for use
only by the individual or entity to which it is addressed
and may contain information that is privileged,
confidential, and exempt from disclosure under applicable
law. If the reader of this message is not the intended
recipient or the employee or agent responsible for
delivering the message to the intended recipient, you are
hereby notified that any dissemination, distribution or
copying of this communication is strictly prohibited. If
you have received this communication in error, please
contact the sender immediately and destroy the material in
its entirety, whether electronic or hard copy. Thank you.
|
