Reconstruction by macro

[ye chen]

Hi All,

I am a new user for quick palm in Melbourne.

I noticed that no matter what parameter I put in the macro for
reconstruction (i.e. 2D particle intensity (16-bit)), the result is a
RGB file. There was however no problem if I did the reconstruction
manually. Is it a bug or is there something I was missing?

Besides, can anyone let me know what is the unit of FWHM in simulate
the sub-diffraction spot option? I found that 100 gave a comparable
result to a simulation of 10nm Guassian by the Zeiss software.

Cheers,


Ye

[Ricardo Henriques]

Hi Ye,

> I noticed that no matter what parameter I put in the macro for
> reconstruction (i.e. 2D particle intensity (16-bit)), the result is a
> RGB file. There was however no problem if I did the reconstruction
> manually. Is it a bug or is there something I was missing?

Yes, this is a bug. I actually don't know if the origin is on my code
or on ImageJ itself. I had in mind making an alternative or improved
reconstruction (and drift correction) method but it may take some
time. If you need a fast fix, you may want to look into making a
reconstruction method yourself. It's quite easy assuming you know how
to make a macro in ImageJ. Here's how:
- after reconstruction save the particles-table as a tiff (this is
just a menu command under Plugins/QuickPALM)
- then open the particles table as an image
- the pixels in the first column are particles intensity, 2nd column
x-position, 3rd column y-position (see a better description here
http://code.google.com/p/quickpalm/wiki/Tutorial_Particle_table)
- generate back image with for example a 10x the size of a camera images
- loop through all the particles and add +1 to the corresponding pixel
to where the particle would be

> Besides, can anyone let me know what is the unit of FWHM in simulate
> the sub-diffraction spot option? I found that 100 gave a comparable
> result to a simulation of 10nm Guassian by the Zeiss software.

That option simply calls an ImageJ gaussian blur after you make the
reconstruction, it's the equivalent to Process\Filters\Gaussian Blur.
See the definition of the Gaussian Blur radius here
http://rsbweb.nih.gov/ij/developer/api/index.html.

Hope it helps.
Cheers,
-R
--
Ricardo Henriques, PhD
Institut Pasteur (Paris, France).
For contact information see: https://sites.google.com/site/paxcalpt/