Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Hello All
I've been toying with these issues a bit of late.

Embedding the beads in optical cements that can have an RIs near 1.51
will certainly help correct spherical problems for oil immersion lenses
but it will make matters a lot worse for water immersion lenses.

You can get optical cements that have RIs ~1.59 similar to the material
of the beads.  However this results in axial distortion due to the
surrounding higher RI even with oil lenses.  It makes imaging really
difficult with water lenses when you embed the beads in something like this.

Ideally you'd probably want to put the beads in something with an RI
similar to the RI of your standard sample (whatever that is) and image
with the appropriate immersion lens.  The only problem is that lower RI
(water-like) immersion media aren't particularly viscous so the beads
move around.

I'd be interested in hearing from anyone with suggestions on a low RI
gel which stays optically clear but is firm enough to stop small beads
moving.  I've tried various PVA mixes but they tend to shrink or cure to
an RI about 1.4-1.45 range.  Matrigel gets awfully scattering when it
sets up.

I haven't tried QDots but in theory this would seem to be an ideal
application but knowing whether you have clusters or individual dots
would be a challenge.

Ian Clements
Imaging Technology
Molecular Probes Inc.



Lutz Schaefer wrote:

>Search the CONFOCAL archive at
>http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
>Hi Peter,
>
> >
> >       Question: is it possible (and has it been tried) to calculate the
> > PSF from a "supra-resolution" target?  If you know the precise dimension
> > of the "ideal sphere", then dividing the Fourier transform of the
>acquired
> > image by the transform of the "ideal sphere" (i.e. deconvolving) should
> > yield the transform of the PSF.  Then the inverse will be the PSF image
> > itself, which can then be used for deconvolution using 3rd party software
> > packages.
> >
>
>Yes, it had been done. The (relatively old) Zeiss KS400 software uses
>relatively large beads (~1 micron or larger depending on objective lens) to
>calculate a measured PSF. An iterative maximum likelihood algorithm is used
>for that. The only disadvantage of the method is that larger beads may not
>contain enough high frequencies to be restored, leaving you with a slightly
>too large restored PSF.
>
>Cheers
>Lutz
>
>____________________________________
>Lutz Schaefer
>Advanced Imaging Methodology Consultation
>16-715 Doon Village Rd.
>Kitchener, Ontario, N2P 2A2
>Website: http://home.golden.net/~lschafer/
>____________________________________
>
>