Re: Neurite tracer

Posted by Fabrice Senger on
URL: http://imagej.273.s1.nabble.com/Neurite-tracer-tp3692383p3692386.html

Gabriel Lapointe a écrit :

> Well I assumed that the roots originate from some center point that is
> big and will produce weird thing in the skeletonization. Beside if the
> center point is still there, the results would be the sum of all the
> roots origination from that point.
>
> So what I meant is that you would have to remove that center point, a
> simple elliptical selection and a clear will do the trick and is faster
> than the whole measurements with the wand ( it might be possible to do
> it automatically but I can't say for sure without seeing an image).
>
> Analyze particle with the "show count mask" option enable will give you
> an numbered outline of objects.
>
> Gabriel
>
> Josh Klein wrote:
>  
>> What do you mean by 'manually remove the seedling'? I could measure the individual seedlings just using the wand tool. I'm trying to avoid treating each seedling separately, but if it comes to that, I'll just use the wand.
>> Is there a program that can outline and label (number) elements on an image?
>> Thanks again,
>> josh
>>
>> --- On Thu, 5/28/09, Gabriel Lapointe <[hidden email]> wrote:
>>
>> From: Gabriel Lapointe <[hidden email]>
>> Subject: Re: Neurite tracer
>> To: [hidden email]
>> Date: Thursday, May 28, 2009, 8:45 AM
>>
>> Hi Josh,
>> Just brainstorming here but, maybe something like this could be a
>> starting point.
>> * Threshold and make the image binary
>> * Manually remove the seedling
>> * run Process > Binary > Skeletonize
>> * run Analyze > Analyze particles.
>>
>> Assuming that the root or not entangled each particle would be a single
>> root length by a single pixel width. I'm not aware of a way to
>> automatically measure the length in ImageJ. however, the length could
>> therefore be *estimated* as either:
>>
>> area/pixel_width  OR  perimeter/2-1pixel_width
>>
>> My understanding of how ImageJ measure the perimeter, I would bet that
>> the second method can give very bad results.
>>
>> Good luck,
>> Gabriel Lapointe
>>
>> Josh Klein wrote:
>>  
>>    
>>> I did look at that. What I am trying to do is measure individual roots on young  seedlings, where there are anywhere from 10-25 seedlings in on image. The roots are not yet all tangled with each other, so I really want a program that identifies individual roots and measures the length. I can do this manually, of course, on the screen using ImageJ measuring wands,  but my students and I both are getting repetitive strain injury! Both neurite tracing programs measure _all_ the neurites at once, and I want automatic definition
>>> of individual entities.
>>> Thanks for your suggestion; any thoughts on the more closely defined problem above?
>>> Cheers,
>>> josh
>>> --- On Thu, 5/28/09, Albert Cardona <[hidden email]> wrote:
>>>
>>>
>>> From: Albert Cardona <[hidden email]>
>>> Subject: Re: Neurite tracer
>>> To: [hidden email]
>>> Date: Thursday, May 28, 2009, 3:40 AM
>>>
>>>
>>> Josh Klein wrote:
>>>    
>>>    
>>>      
>>>> I am trying to use the Neurite Tracer plugin. When I get to the step
>>>> for "Open images for setting threshold", the program won't let me
>>>> select images from a file. It only recognizes the folders themselves,
>>>> which of course are not .tif images that the plugin requires, and then it tosses me out of the
>>>> program.
>>>> Suggestions leading to success would be appreciated.
>>>>    
>>>>      
>>>>      
>>>>        
>>> Josh,
>>>
>>> Did you have a look at the "Simple Neurite Tracer" by Mark Longair?
>>>
>>> http://pacific.mpi-cbg.de/wiki/index.php/Simple_Neurite_Tracer
>>> http://homepages.inf.ed.ac.uk/s9808248/imagej/tracer/
>>>
>>>
>>> Albert
>>>
>>> -- Albert Cardona
>>> http://albert.rierol.net
>>>
>>>
>>>
>>>        
>>>
>>>    
>>>    
>>>      
>>
>>      
>>
>>  
>>    
Hi there,

indeed it would make sens to see some of your images. The problem
reminds me vessel outgrowth from aortic rings or neurite outgrowth as
observed in pc12 cells.
There are some semi automatic procedures but as stated by Albert, the
solution might be dedicated to your particular needs.

Fabrice.

--
Senger Fabrice