Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
The pinhole diameter is based upon the the diameter of the Airy disk
at the detector. The usual definition of optimal pinhole diameter is
that the confocal aperture is set to the outside edge of the first
dark ring of the Airy disk. This diameter depends upon the
Magnification, emission wavelength and distance between the back
focal plane of the objective and the detector. The table in your
1024 manual has converted measurements in optical units to mm of
pinhole diameter. The 1.3 meter folded optical path in the scanhead
allowed these systems to avoid the issues of pinhole alignment that
plagued other instruments of that era. The table assumes you are
using a Nikon 160 mm tube length microscope, such as a Diaphot.
Notice that the equation gives two examples, a standard brightfield
scope or with addition of fluorescence and DIC. If you only have epi-
fluorescence on your scope, and no DIC, then use assume an total mag
of 63 times lens mag and use the average of the 2 values given in the
table for your lens.
Nyquist sampling at 2.3 times the object frequency is adequate for
detection, but to resolve the point to point Rayleigh resolution, you
need to double it again to 4.6X the object frequency, especially if
deconvolving.
Regards,
glen
Glen MacDonald
Core for Communication Research
Virginia Merrill Bloedel Hearing Research Center
Box 357923
University of Washington
Seattle, WA 98195-7923 USA
(206) 616-4156
[log in to unmask]
************************************************************************
******
The box said "Requires WindowsXP or better", so I bought a Macintosh.
************************************************************************
******
On Feb 9, 2008, at 3:30 PM, Pedro J Camello wrote:
> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
> Yes, nm.
>
> Julio, you are right about lateral resolution. My reasoning is that
> although the optical performance does not depend on the structure,
> 150 nm
> is supposed to be good enough to sample structures of 500 nm. I
> would like
> to avoid too much oversampling.
>
> I don´t know how Biorad calculated that 2.6 zoom is the maximum
> advisable
> for this objective (it is a shame, but I never worried before about
> sampling rate, resolution, etc)
>
> Many thanks for your help.
>
>> Search the CONFOCAL archive at
>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>>
>> Hello All,
>>
>> To avoid more confusion, you both meant nanometers, not
>> micrometers, when
>> talking about xy-, z-resolution and z-stepsize, correct?
>> E.g., "optimal" xy resolution is 200 nm, not microns, with pixel
>> size at
>> 80
>> nm with Nyquist, 160 nm z-stepsize, etc.
>>
>> Zoltan
>>
>>
>> On 2/9/08, Julio Vazquez <[log in to unmask]> wrote:
>>>
>>> Search the CONFOCAL archive at
>>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal = Hi Pedro,
>>>
>>>
>>> I am not sure how you or Biorad are calculating the optimal pixel
>>> sizes,
>>> or what the specific constraints for the Biorad 1024 are. For a
>>> 60x/1.4NA
>>> objective, the theoretical resolution in x/y at 580 nm (if you
>>> are using
>>> a
>>> red mito tracker) is around 200 microns (I am using the formula d
>>> = 0.5
>>> x
>>> lambda/ NA). Therefore, your optimal pixel size, if you use a
>>> Nyquist
>>> factor
>>> of 2.5, should be around 80 microns. Obviously, you may get
>>> slightly
>>> different values depending on how your choice of wavelength.
>>>
>>>
>>> Regarding the sampling along the z axis, it is not so much the
>>> dimensions
>>> of the structures (mitochondria) that matter, but the optical
>>> properties
>>> of
>>> your scope. The axial resolution is about half the x/y
>>> resolution, and
>>> therefore 160 micron spacing between sections would be about
>>> optimal.
>>> If
>>> you are trying to look at very fine structures, oversampling a
>>> little
>>> (both
>>> in x, y, and z) may not be a bad idea, as long as your sample can
>>> take
>>> it
>>> and doesn't bleach...
>>>
>>>
>>> I don't think it as an issue that your x/y pixel dimensions are
>>> different
>>> from your z section spacing. The PSF itself is not spherically
>>> symmetrical
>>> (but elongated along z) and your sampling along x/y and z should
>>> be set
>>> accordingly. Secondly, when you load your images into whatever
>>> program
>>> you
>>> use for doing the 3-D measurements, the image should be
>>> calibrated to
>>> reflect the actual x,y and z dimensions. As long as you used Nyquist
>>> sampling, and you have good signal-to-noise, you should be able
>>> to get
>>> accurate measurements.
>>>
>>>
>>> For deconvolution, we've had good results with both measured PSF and
>>> calculated PSFs; blind should work fine too, but I haven't tried.
>>> If you
>>> have some reference beads, you could image them under conditions
>>> similar
>>> to
>>> those used to image your sample, deconvolve them with the same
>>> method,
>>> and
>>> do some measurements. That should give you an idea of how well your
>>> deconvolution alrgorithm is performing. If it doesn't look good,
>>> you can
>>> always re-deconvolve your data with a different method.
>>>
>>>
>>>
>>>
>>>
>>>
>>> Julio.
>>>
>>>
>>>
>>> --
>>> Julio Vazquez, PhD
>>> Director of Scientific Imaging
>>> Fred Hutchinson Cancer Research Center
>>> 1100 Fairview Ave. N., mailstop DE-512
>>> Seattle, WA 98109-1024
>>>
>>>
>>> Tel: Office: 206-667-1215/ Lab: 206-667-4205
>>> FAX: 206-667-6845
>>>
>>>
>>> [log in to unmask]
>>> http://www.fhcrc.org/science/shared_resources/imaging/
>>>
>>>
>>>
>>> --------------------------------------------------
>>> *DISCLAIMER:*
>>>
>>> This message is confidential, intended only for the named
>>> recipient(s)
>>> and
>>> may contain information that is privileged or exempt from disclosure
>>> under
>>> applicable law. If you are not the intended recipient(s), you are
>>> notified
>>> that the dissemination, distribution or copying of this
>>> information is
>>> strictly prohibited. If you received this message in error, please
>>> notify
>>> the sender then delete this message.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Feb 9, 2008, at 9:30 AM, Pedro J Camello wrote:
>>>
>>> Search the CONFOCAL archive at
>>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>>>
>>>
>>> I´m trying to image a 3D map of mitotracker-stained cells. I´m
>>> using a
>>> Biorad 1024, 60x, NA 1.4. My first attempts generated acceptable
>>> images
>>> with iris 1.6 -2 (close to the theoretical optimal iris indicated by
>>> Biorad with this objective).
>>>
>>>
>>> My 2 questions are:
>>>
>>>
>>> 1) Using zoom 2.6 (the theoretical optimum zoom for this
>>> objective) I
>>> get
>>> pixel size 166 µm. Since published images with my cell type
>>> (pancreatic
>>> acini) indicate that most of the mitochondria are around 500 nm
>>> diameter
>>> (similar to the measured resolution of mi micro with iris 2) my
>>> intend
>>> is
>>> to sample at 150 nm to meet Nyquist. However, this creates a non-
>>> cubic
>>> voxel (166x166x150). Will be this a problem to make posterior 3D
>>> visualization and volume measurements?
>>>
>>>
>>> 2) If I need deconvolution to posterior processing and
>>> measurements, is
>>> it
>>> essential PSF measurements or will blinded methods be enough?
>>>
>>>
>>> Thanks in advance
>>>
>>>
>>> --
>>> Dr Pedro J Camello
>>> Dpt Physiology
>>> Faculty of Veterinary Sciences
>>> University of Extremadura
>>> 10071 Caceres
>>> Spain
>>> Ph: 927257100 Extension 1321/1290
>>> Fax:927257110
>>>
>>>
>>>
>>
>>
>> --
>> --
>> Zoltan Cseresnyes
>> Facility manager, Imaging Suite
>> Dept. of Zoology University of Cambridge
>> Downing Street, Cambridge
>> CB2 3EJ UK
>>
>> Tel.: (++44) (0)1223 769282
>> Fax.: (++44) (0)1223 336676
>>
>
>
> --
> Dr Pedro J Camello
> Dpt Physiology
> Faculty of Veterinary Sciences
> University of Extremadura
> 10071 Caceres
> Spain
> Ph: 927257100 Extension 1321
> Fax:927257110
|
