quantifying tortuosity of blood vessels in 3d

[Romin, Yevgeniy/Sloan Kettering Institute]

Dear List

 

I am currently working on an image analysis project that involves looking at 3d z-stacks of a network of blood vessels.  Most of these blood vessels are smooth and regular, but there are small segments of them that are distorted, bent and twisted.  We need to quantify this “twistedness” in these segments and compare it to the regular segments.  I am wondering if any of you have any experience doing this sort of analysis.  This involves quantifying tortuosity of the vessels and their distortion from the regular pattern.  The papers which I found on the subject use medical software specifically designed for this function, and I have no access to them.  Some of the things I tried include using the angiogenesis tube formation module in metamorph, counting the “nodes” or “branch points” as well as other functions.  I do have access to Metamorph, Volocity, ImageJ,  Axiovision and Imaris.  The other thing that would be preferable is to do this sort of analysis in 3d, on the reconstructed z-stacks instead of going plane by plane.  If anybody has any experience with this or a similar analysis, any input would be greatly appreciated.

 

Thanks very much to everybody in advance,

 

---------------------------------------------------

Yevgeniy Romin

 

Digital Microscopist

Memorial Sloan-Kettering Cancer Center

Molecular Cytology Core Facility

1275 York Ave. Box 333

New York, NY 10065

Tel.[phone]

Fax. [phone]

---------------------------------------------------

 

 
     =====================================================================
     
     Please note that this e-mail and any files transmitted with it may be 
     privileged, confidential, and protected from disclosure under 
     applicable law. If the reader of this message is not the intended 
     recipient, or an employee or agent responsible for delivering this 
     message to the intended recipient, you are hereby notified that any 
     reading, dissemination, distribution, copying, or other use of this 
     communication or any of its attachments is strictly prohibited.  If 
     you have received this communication in error, please notify the 
     sender immediately by replying to this message and deleting this 
     message, any attachments, and all copies and backups from your 
     computer.

[Martin Wessendorf-2]

Dear Yevgeniy--

Yevgeniy Romin wrote:

> I am currently working on an image analysis project that involves
> looking at 3d z-stacks of a network of blood vessels.  Most of these
> blood vessels are smooth and regular, but there are small segments of
> them that are distorted, bent and twisted.  We need to quantify this
> “twistedness” in these segments and compare it to the regular segments.  
> I am wondering if any of you have any experience doing this sort of
> analysis.  This involves quantifying tortuosity of the vessels and their
> distortion from the regular pattern.  The papers which I found on the
> subject use medical software specifically designed for this function,
> and I have no access to them.  Some of the things I tried include using
> the angiogenesis tube formation module in metamorph, counting the
> “nodes” or “branch points” as well as other functions.  I do have access
> to Metamorph, Volocity, ImageJ,  Axiovision and Imaris.  The other thing
> that would be preferable is to do this sort of analysis in 3d, on the
> reconstructed z-stacks instead of going plane by plane.  If anybody has
> any experience with this or a similar analysis, any input would be
> greatly appreciated.

Seems as if the first question would be how you want to define
"tortuosity".  If you're interested solely in how much blood vessel is
packed into a given region, that could be defined simply as length per
unit volume.  However, it sounds as if you're interested in changes in
direction and branching as well, and that would be more complicated.

Whatever you're interested in, I'd suggest that you consult a good book
on unbiased stereology, such as "Unbiased Stereology: Three-Dimensional
Measurement in Microscopy" by CV Howard and M Reed, or "Principles and
Practices of Unbiased Stereology: An Introduction for Bioscientists" by
PR Mouton.  There are a lot of ways to get incorrect answers when
addressing questions like these, so you may want to spend some time with
the books before you work with the software.

Good luck!

Martin Wessendorf

--
Martin Wessendorf, Ph.D.                   office: [phone]
Assoc Prof, Dept Neuroscience                 lab: [phone]
University of Minnesota             Preferred FAX: [phone]
6-145 Jackson Hall, 321 Church St. SE    Dept Fax: [phone]
Minneapolis, MN  55455                    e-mail: [hidden email]

[Glen MacDonald-2]

Dear Yevgeniy,
We've started looking at something similar with glial processes and neurites.  An initial approach that seems promising uses the Simple Neurite Tracer plugin bundled with FIJI to detect and trace linear structures through a stack.  the plugin delivers the length of the trace and the x,y,z coordinates of the beginning and ending points to allow computation of the straight line distance.  

Regards,
Glen



On Apr 28, 2010, at 5:48 AM, Yevgeniy Romin wrote:

> Dear List
>  
> I am currently working on an image analysis project that involves looking at 3d z-stacks of a network of blood vessels.  Most of these blood vessels are smooth and regular, but there are small segments of them that are distorted, bent and twisted.  We need to quantify this “twistedness” in these segments and compare it to the regular segments.  I am wondering if any of you have any experience doing this sort of analysis.  This involves quantifying tortuosity of the vessels and their distortion from the regular pattern.  The papers which I found on the subject use medical software specifically designed for this function, and I have no access to them.  Some of the things I tried include using the angiogenesis tube formation module in metamorph, counting the “nodes” or “branch points” as well as other functions.  I do have access to Metamorph, Volocity, ImageJ,  Axiovision and Imaris.  The other thing that would be preferable is to do this sort of analysis in 3d, on the reconstructed z-stacks instead of going plane by plane.  If anybody has any experience with this or a similar analysis, any input would be greatly appreciated.
>  
> Thanks very much to everybody in advance,
>  
> ---------------------------------------------------
> Yevgeniy Romin
>  
> Digital Microscopist
> Memorial Sloan-Kettering Cancer Center
> Molecular Cytology Core Facility
> 1275 York Ave. Box 333
> New York, NY 10065
> Tel.[phone]
> Fax. [phone]
> ---------------------------------------------------
>

Glen MacDonald
Core for Communication Research
Virginia Merrill Bloedel Hearing Research Center
Box 357923
University of Washington
Seattle, WA 98195-7923  USA
[phone]
[hidden email]

[Jerry (Gerald) Sedgewick]

Yevgeniy,

I've calculated major vs. minor axis of tortuous vessels in 2D using
FIJI and this worked very well as an indicator of tortuosity.  I created
smaller discrete segments of vessels using a grid as a mask, and then
removed all partial lengths via the perimeter values.   I don't know
offhand how this could be implemented in 3D, but it was a fairly easy
way to get results.

Another method I've come across uses a straight line (as mentioned by
Glen) and then measures angles from that line, but not with off the
shelf software.  This method also separates vessels into smaller lengths
before analysis.  I would think that if a straight line could be drawn
between end points, and then that could be ratioed against the perimeter
of a skeletonized vessel, degrees of tortuousity could be discovered.

With stereology, one would think that a tortuous vessel would have a
greater likelihood of taking up more volume, and so the Cavalerie probe
might work, but I found that the frequency of the crosshairs had to be
so dense (I never did find out how dense before abandoning stereology)
that the efficiency of this measurement tanks.  But you might want to
try this.

The ability to detect tortuousity is important in the ophthalmology
world, especially in the automated detection of diabetic retinopathy.  
Maybe you would have success finding papers in this neck of the woods.

Best,
Jerry

Glen MacDonald wrote:

> Dear Yevgeniy,
> We've started looking at something similar with glial processes and neurites.  An initial approach that seems promising uses the Simple Neurite Tracer plugin bundled with FIJI to detect and trace linear structures through a stack.  the plugin delivers the length of the trace and the x,y,z coordinates of the beginning and ending points to allow computation of the straight line distance.  
>
> Regards,
> Glen
>
>
>
> On Apr 28, 2010, at 5:48 AM, Yevgeniy Romin wrote:
>
>  
>> Dear List
>>  
>> I am currently working on an image analysis project that involves looking at 3d z-stacks of a network of blood vessels.  Most of these blood vessels are smooth and regular, but there are small segments of them that are distorted, bent and twisted.  We need to quantify this “twistedness” in these segments and compare it to the regular segments.  I am wondering if any of you have any experience doing this sort of analysis.  This involves quantifying tortuosity of the vessels and their distortion from the regular pattern.  The papers which I found on the subject use medical software specifically designed for this function, and I have no access to them.  Some of the things I tried include using the angiogenesis tube formation module in metamorph, counting the “nodes” or “branch points” as well as other functions.  I do have access to Metamorph, Volocity, ImageJ,  Axiovision and Imaris.  The other thing that would be preferable is to do this sort of analysis in 3d, on the reconstructed z-stacks instead of going plane by plane.  If anybody has any experience with this or a similar analysis, any input would be greatly appreciated.
>>  
>> Thanks very much to everybody in advance,
>>  
>> ---------------------------------------------------
>> Yevgeniy Romin
>>  
>> Digital Microscopist
>> Memorial Sloan-Kettering Cancer Center
>> Molecular Cytology Core Facility
>> 1275 York Ave. Box 333
>> New York, NY 10065
>> Tel.[phone]
>> Fax. [phone]
>> ---------------------------------------------------
>>
>>    
>
>
>
> Glen MacDonald
> Core for Communication Research
> Virginia Merrill Bloedel Hearing Research Center
> Box 357923
> University of Washington
> Seattle, WA 98195-7923  USA
> [phone]
> [hidden email]
>
>  

--
Jerry (Gerald) Sedgewick
Author: "Scientific Imaging with Photoshop: Methods, Measurement and Output."

Sedgewick Initiatives
965 Cromwell Avenue
Saint Paul, MN  55114
651-788-2261
[hidden email]
http://www.quickphotoshop.com
http://www.rawlight.com
http://www.jerrysedgewick.com