Nearest Neighbor Analysis

[Matt Swulius]

Hello all,
I am currently doing spatial analysis of gold markers in electron
micrographs. I would like to do nearest neighbor analysis, which seems
fairly straightforward conceptually, but have not been able to find a
utility that will actually calculate the nearest neighbor distances.  Does
anyone have experience doing this or know if imageJ is even capable of doing
this? I've looked through the plugins on their site and couldn't find
anything for doing this specifically, but it's possible I overlooked
something.

Thanks in advance for any information you can provide.

Matt Swulius

PhD Student
Dept. of Neurobiology and Anatomy
UTHSC - Houston

[dpoburko]

Matt Swulius wrote:

> Hello all,
> I am currently doing spatial analysis of gold markers in electron
> micrographs. I would like to do nearest neighbor analysis, which seems
> fairly straightforward conceptually, but have not been able to find a
> utility that will actually calculate the nearest neighbor distances.  Does
> anyone have experience doing this or know if imageJ is even capable of doing
> this? I've looked through the plugins on their site and couldn't find
> anything for doing this specifically, but it's possible I overlooked
> something.
>
> Thanks in advance for any information you can provide.
>
> Matt Swulius
>
> PhD Student
> Dept. of Neurobiology and Anatomy
> UTHSC - Houston
>  

Hey Matt,

  If you are only looking at 2D images, then an easy method would be to
use a euclidean distance map. 1. Threshold your image to isolate regions
over your gold particles. You could erode these regions to a centroid to
"increase precision." 2. Generate the distance map (built into IJ i
believe). By definition, this will create an image with intensity values
equal to the distance between objects.  Now dilate you regions by a
pixel or two to measure the distance between regions of interest (i.e.
gold particles). If you measured the minimum pixel value surrounding you
region on the distance map, this should be your distance to the nearest
neighbouring region/particle. It's not the most elegant way to go about
it, but it's doable. Alternatively, you could just export the centroid
co-ordinates of your gold particles (again, thresholded rgeions), and
calculate the nearest neighbour distance in excel.

Cheers,
Damon



--

Damon Poburko, PhD
Postdoctoral Research Fellow
Stanford University School of Medicine
Dept. of Molecular & Cellular Physiology
279 Campus Dr., Beckman B103, Stanford, CA 94305
Ph: 650 725 7564, fax: [phone]