Reconstructing a 3D image of an object in water

[sjhenn]

Dear List,

I have a question which I hope someone can help me with;

We have an MRI scan of some coral in a tube full of seawater. We are trying to image this coral in 3D (removing the tube and the water), but even through the 3D imaging plugins, we can only ever get a 3D image of the tube that the coral in is. We have tried subtracting the background but this only got rid of the water within the tube, and not the tube itself. We can visualise the coral in various ways but seem unable to isolate it. If anyone has any advice on this it would be greatly appreciated!

Best wishes,
Seb

[Michael Doube-3]

Hi Seb,

On 19.07.12 10:11 PM, sjhenn wrote:
> we can only ever get a 3D image of the tube
> that the coral in is. We have tried subtracting the background but this only
> got rid of the water within the tube, and not the tube itself. We can
> visualise the coral in various ways but seem unable to isolate it.

It would be really helpful if you could post an image somewhere so that
we can see your problem. A single slice showing the pixel values
representing air, water, tube and coral would be adequate. Also, is it a
hard or soft coral, and what MRI sequence did you use?

Michael

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[sjhenn]

<quote author="Michael Doube-3">
Hi Seb,

It would be really helpful if you could post an image somewhere so that
we can see your problem. A single slice showing the pixel values
representing air, water, tube and coral would be adequate. Also, is it a
hard or soft coral, and what MRI sequence did you use?

It is hard calcium carbonate - Attached are screenshots of the 2dseq file opened as is (where the water is orange/red) with the 3D view next to it. The second screenshot is once converted to binary to make the water at least all white but the same problem of isolating the coral within the tube. We are planning on using this image for 3D printing of the coral. Any advice greatly appreciated!

[Michael Doube-3]

> It is hard calcium carbonate

Do you have access to a CT scanner? You will get much better contrast
from the calcium carbonate in that modality than MRI, which is typically
poor at giving contrast from mineralised tissues (unless you find a
specialized sequence).

> - Attached are screenshots of the 2dseq file
> opened as is (where the water is orange/red) with the 3D view next to it.
> The second screenshot is once converted to binary to make the water at least
> all white but the same problem of isolating the coral within the tube.

If you have to use this image, you need to make a mask which describes
the total tube volume (fill holes might work), then AND it with an
inverted version of the tube+coral, to get a result which is just the coral.

Michael

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Julian Cooper]

Dear Seb,

Would it be possible for you to create a circular ROI around the coral, then
"Edit>Clear Outside" (all slices) to remove the tube before reconstructing?
That would reduce the problem to differentiating between the coral and
water. As Michael has said that would be much easier on CT than MRI.

I often meet similar issues when 3D reconstructing CT scans of the pelvis.
There are typically metal structures (stretcher, external fixators,
monitoring equipment etc) that lie around the patient and need to be
isolated from the bone to get a decent reconstruction-the "clear outside"
method works very well for that.

Kind regards,

Julian

Julian Cooper
Consultant Orthopaedic Trauma Surgeon
Birmingham UK

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of sjhenn
Sent: 19 July 2012 21:11
To: [hidden email]
Subject: Reconstructing a 3D image of an object in water

Dear List,

I have a question which I hope someone can help me with;

We have an MRI scan of some coral in a tube full of seawater. We are trying
to image this coral in 3D (removing the tube and the water), but even
through the 3D imaging plugins, we can only ever get a 3D image of the tube
that the coral in is. We have tried subtracting the background but this only
got rid of the water within the tube, and not the tube itself. We can
visualise the coral in various ways but seem unable to isolate it. If anyone
has any advice on this it would be greatly appreciated!

Best wishes,
Seb

--
View this message in context:
http://imagej.1557.n6.nabble.com/Reconstructing-a-3D-image-of-an-object-in-w
ater-tp4999503.html
Sent from the ImageJ mailing list archive at Nabble.com.

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[sjhenn]

<quote author="Julian Cooper">
Dear Seb,

Would it be possible for you to create a circular ROI around the coral, then
"Edit>Clear Outside" (all slices) to remove the tube before reconstructing?
That would reduce the problem to differentiating between the coral and
water. As Michael has said that would be much easier on CT than MRI.

I often meet similar issues when 3D reconstructing CT scans of the pelvis.
There are typically metal structures (stretcher, external fixators,
monitoring equipment etc) that lie around the patient and need to be
isolated from the bone to get a decent reconstruction-the "clear outside"
method works very well for that.

Kind regards,

Julian

Dear Michael and Julian,

Thank you for your advice - I do not have access to a CT system so I have to stick to the MRI image. Using a combination of converting the image to binary, selecting the region of interest, and inverting the image of everything outside of this I managed to get the image isolated. Now I can work on tidying up this technique to get a cleaner image. Thanks again,
Seb