Deconvolution of 3D SIM data

[Martin Wessendorf-2]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

On 4/13/2011 5:28 PM, Guy Cox wrote:

> Abbe considered rays diffracted by two points on the sample.  The points
> will be resolved if their diffracted rays can enter the objective.  This
> can only apply to  a specimen illuminated from an external source.  In
> fluorescence each point emits light and this is totally incoherent with
> respect to light from another point.   There is no diffraction at the
> sample so Abbe's calculation cannot be applied.  Rayleigh's criterion,
> based on how the optics turn the image of a point into a disk (the Airy
> disk) does apply.

This is (for me!) a very intuitive explanation, but it suggests that
with fluorescence, arbitrarily small resolution can be attained given
sufficiently high s/n.  That sounds something like what you said in your
earlier posting, except for the phrase "arbitrarily small".

Is that correct?  If not, what is the absolute limit of resolution in
fluorescence?

Martin

--
Martin Wessendorf, Ph.D.                   office: [phone]
Assoc Prof, Dept Neuroscience                 lab: [phone]
University of Minnesota             Preferred FAX: [phone]
6-145 Jackson Hall, 321 Church St. SE    Dept Fax: [phone]
Minneapolis, MN  55455                    e-mail: [hidden email]

[Andreas Bruckbauer]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

I think this is the conclusion of a paper by Sripad Ram (PNAS 103, 2006, 4457), but i think in this case you have to know the number of objects you are looking at.

best wishes

Andreas

 

 


 

 

-----Original Message-----
From: Martin Wessendorf <[hidden email]>
To: [hidden email]
Sent: Thu, 14 Apr 2011 4:41
Subject: Re: Deconvolution of 3D SIM data


*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy 
*****
 
On 4/13/2011 5:28 PM, Guy Cox wrote:
 
> Abbe considered rays diffracted by two points on the sample.  The points
> will be resolved if their diffracted rays can enter the objective.  This
> can only apply to  a specimen illuminated from an external source.  In
> fluorescence each point emits light and this is totally incoherent with
> respect to light from another point.   There is no diffraction at the
> sample so Abbe's calculation cannot be applied.  Rayleigh's criterion,
> based on how the optics turn the image of a point into a disk (the Airy
> disk) does apply.
 
This is (for me!) a very intuitive explanation, but it suggests that with fluorescence, arbitrarily small resolution can be attained given sufficiently high s/n.  That sounds something like what you said in your earlier posting, except for the phrase "arbitrarily small".
 
Is that correct?  If not, what is the absolute limit of resolution in fluorescence?
 
Martin
 
-- Martin Wessendorf, Ph.D.                   office: [phone]
Assoc Prof, Dept Neuroscience                 lab: [phone]
University of Minnesota             Preferred FAX: [phone]
6-145 Jackson Hall, 321 Church St. SE    Dept Fax: [phone]
Minneapolis, MN  55455                    e-mail: [hidden email]

 

[Greg Martin-8]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hey Folks --

    Great discussion.  To answer Martin's question as to what ultimately limits resolution in fluorescence I would say: The specimen!  How many of us have specimens with structural preservation at the level of 20 nm, or even 50 nm?  Do we know what that preservation really is at 20 nm -- what it "should" look like?  All these valuable discussions considering resolution are assuming an a priori  knowledge of the specimen we don't have when it comes to biology.  This is a good thing.  Once we know what the new systems can do they can be applied to our unknowns -- like the cells we've been examining by fluorescence for so many years at 200 nm resolution.  Should be fun!

Be peace!  Greg.

Greg Martin

Keck Microscopy Facility
University of Washington School of Medicine
www.depts.washington.edu/keck

206-685-8784 (office)
425-344-2632 (cell)
----- Original Message -----
From: "Martin Wessendorf" <[hidden email]>
To: <[hidden email]>
Sent: Wednesday, April 13, 2011 8:41 PM
Subject: Re: Deconvolution of 3D SIM data

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> On 4/13/2011 5:28 PM, Guy Cox wrote:
>
>> Abbe considered rays diffracted by two points on the sample.  The points
>> will be resolved if their diffracted rays can enter the objective.  This
>> can only apply to  a specimen illuminated from an external source.  In
>> fluorescence each point emits light and this is totally incoherent with
>> respect to light from another point.   There is no diffraction at the
>> sample so Abbe's calculation cannot be applied.  Rayleigh's criterion,
>> based on how the optics turn the image of a point into a disk (the Airy
>> disk) does apply.
>
> This is (for me!) a very intuitive explanation, but it suggests that
> with fluorescence, arbitrarily small resolution can be attained given
> sufficiently high s/n.  That sounds something like what you said in your
> earlier posting, except for the phrase "arbitrarily small".
>
> Is that correct?  If not, what is the absolute limit of resolution in
> fluorescence?
>
> Martin
>
> --
> Martin Wessendorf, Ph.D.                   office: [phone]
> Assoc Prof, Dept Neuroscience                 lab: [phone]
> University of Minnesota             Preferred FAX: [phone]
> 6-145 Jackson Hall, 321 Church St. SE    Dept Fax: [phone]
> Minneapolis, MN  55455                    e-mail: [hidden email]
>

[Steffen Dietzel]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi everybody,

somewhat related to the ongoing discussion on resolution, I came across
a puzzle today concerning the resolution of multi-photon microscopy.

I measured the resolution of our third harmonic generation (THG)
microscope and surprisingly I came up with a full width half maximum
(FWHM) slightly better than theory allows. Seems the most likely
explanation is I applied the wrong theory. But which one is the correct one?

The experiment:
THG with 1275 nm, Objective 0.95 NA (water, 20x), beads 60 nm in 2%
agarose, voxel size 0.136 x 0.136 x 0.5 µm. Result: FWHM ~0.7 µm (for
both, forward and backward THG)

Assuming that for multi-photon point-scanners only the excitation
wavelength is relevant, I used 1275 nm for the theory (Rayleigh):
r=0.61λ/NA = 0.82 µm

So, the measured resolution is one pixel better than the theoretical
limit. You don't get that lucky every day ;-)

Possibilities I have considered:
- I messed up the experiment. I wouldn't know, however, how I could get
a better result by messing up.
- Microscope settings are wrong (wrong pixel size). Possible of course
but not very likely.
- Rayleigh does not apply to multi-photon, I overlooked something. If
so, please help out.
- THG requires 3 photons to take place. Maybe the photon density is low
enough in the outer areas of the PSF so that signal generation is
limited to inner areas of the PSF? (Now that would be really interesting
from an academic point of view, since it would mean you could squeeze
the size of the excitation spot relative to the wavelength with 4, 5,
etc. photon effects. Although it probably wouldn't do much good for
practical purposes since you would have to start with long wavelengths
to end up with a visible (=easy detectable) signal.)


Any ideas? Could people share measured FWHMs from their multi-photon
setup? Maybe even from a 3 photon process?

Steffen


--
------------------------------------------------------------
Steffen Dietzel, PD Dr. rer. nat
Ludwig-Maximilians-Universität München
Walter-Brendel-Zentrum für experimentelle Medizin (WBex)
Head of light microscopy

Mail room:
Marchioninistr. 15, D-81377 München

Building location:
Marchioninistr. 27,  München-Großhadern

[Mark Cannell]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

I'm afraid the idea is MUCH older than 2008. I recall reading a paper  
from the (?) 1960's that discussed the idea that  the diffraction  
limit was not a real limit at all. I believe they were using bacteria  
as an example of where knowledge of the object imparts more  
information to overcome the 'limit'

Cheers Mark


On 15/04/2011, at 2:16 AM, Andreas Bruckbauer wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> I think this is the conclusion of a paper by Sripad Ram (PNAS 103,  
> 2006, 4457), but i think in this case you have to know the number of  
> objects you are looking at.
>
> best wishes
>
> Andreas
>
>
>
>
>
>
>
>
>
>
> -----Original Message-----
> From: Martin Wessendorf <[hidden email]>
> To: [hidden email]
> Sent: Thu, 14 Apr 2011 4:41
> Subject: Re: Deconvolution of 3D SIM data
>
>
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> On 4/13/2011 5:28 PM, Guy Cox wrote:
>
>> Abbe considered rays diffracted by two points on the sample.  The  
>> points
>> will be resolved if their diffracted rays can enter the objective.  
>> This
>> can only apply to  a specimen illuminated from an external source.  
>> In
>> fluorescence each point emits light and this is totally incoherent  
>> with
>> respect to light from another point.   There is no diffraction at the
>> sample so Abbe's calculation cannot be applied.  Rayleigh's  
>> criterion,
>> based on how the optics turn the image of a point into a disk (the  
>> Airy
>> disk) does apply.
>
> This is (for me!) a very intuitive explanation, but it suggests that  
> with fluorescence, arbitrarily small resolution can be attained  
> given sufficiently high s/n.  That sounds something like what you  
> said in your earlier posting, except for the phrase "arbitrarily  
> small".
>
> Is that correct?  If not, what is the absolute limit of resolution  
> in fluorescence?
>
> Martin
>
> -- Martin Wessendorf, Ph.D.                   office: [phone]
> Assoc Prof, Dept Neuroscience                 lab: [phone]
> University of Minnesota             Preferred FAX: [phone]
> 6-145 Jackson Hall, 321 Church St. SE    Dept Fax: [phone]
> Minneapolis, MN  55455                    e-mail: [hidden email]
>
>

[Mark Cannell]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

It's multiphoton excitation so the effective PSF is much smaller than  
the 1 photon PSF. This has been described in numerous texts. Not  
wanting to rain on your parade but for THG the resolution should be  
better than you measured.

Cheers



On 15/04/2011, at 5:52 AM, Steffen Dietzel wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hi everybody,
>
> somewhat related to the ongoing discussion on resolution, I came  
> across a puzzle today concerning the resolution of multi-photon  
> microscopy.
>
> I measured the resolution of our third harmonic generation (THG)  
> microscope and surprisingly I came up with a full width half maximum  
> (FWHM) slightly better than theory allows. Seems the most likely  
> explanation is I applied the wrong theory. But which one is the  
> correct one?
>
> The experiment:
> THG with 1275 nm, Objective 0.95 NA (water, 20x), beads 60 nm in 2%  
> agarose, voxel size 0.136 x 0.136 x 0.5 µm. Result: FWHM ~0.7 µm  
> (for both, forward and backward THG)
>
> Assuming that for multi-photon point-scanners only the excitation  
> wavelength is relevant, I used 1275 nm for the theory (Rayleigh):  
> r=0.61λ/NA = 0.82 µm
>
> So, the measured resolution is one pixel better than the theoretical  
> limit. You don't get that lucky every day ;-)
>
> Possibilities I have considered:
> - I messed up the experiment. I wouldn't know, however, how I could  
> get a better result by messing up.
> - Microscope settings are wrong (wrong pixel size). Possible of  
> course but not very likely.
> - Rayleigh does not apply to multi-photon, I overlooked something.  
> If so, please help out.
> - THG requires 3 photons to take place. Maybe the photon density is  
> low enough in the outer areas of the PSF so that signal generation  
> is limited to inner areas of the PSF? (Now that would be really  
> interesting from an academic point of view, since it would mean you  
> could squeeze the size of the excitation spot relative to the  
> wavelength with 4, 5, etc. photon effects. Although it probably  
> wouldn't do much good for practical purposes since you would have to  
> start with long wavelengths to end up with a visible (=easy  
> detectable) signal.)
>
>
> Any ideas? Could people share measured FWHMs from their multi-photon  
> setup? Maybe even from a 3 photon process?
>
> Steffen
>
>
> --
> ------------------------------------------------------------
> Steffen Dietzel, PD Dr. rer. nat
> Ludwig-Maximilians-Universität München
> Walter-Brendel-Zentrum für experimentelle Medizin (WBex)
> Head of light microscopy
>
> Mail room:
> Marchioninistr. 15, D-81377 München
>
> Building location:
> Marchioninistr. 27,  München-Großhadern

[Martin Wessendorf-2]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

On 4/14/2011 3:31 PM, Mark Cannell wrote:

> I'm afraid the idea is MUCH older than 2008. I recall reading a paper
> from the (?) 1960's that discussed the idea that the diffraction limit
> was not a real limit at all. I believe they were using bacteria as an
> example of where knowledge of the object imparts more information to
> overcome the 'limit'

I think no one would deny that if you know what you're looking at, you
can break the "resolution barrier"--STORM and PALM being prime examples.
  However, I'd still be interested in hearing whether--without having
any additional information about the sample--there is any theoretical
limit to resolution in fluorescence microscopy, other than that imposed
by signal-to-noise ratio.

Martin
--
Martin Wessendorf, Ph.D.                   office: [phone]
Assoc Prof, Dept Neuroscience                 lab: [phone]
University of Minnesota             Preferred FAX: [phone]
6-145 Jackson Hall, 321 Church St. SE    Dept Fax: [phone]
Minneapolis, MN  55455                    e-mail: [hidden email]

[Mark Cannell]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

To answer your question directly, there is no information beyond the  
highest spatial frequency contained in the PSF.  The power spectrum  
does not fall to zero at the diffraction limit but rolls off very  
rapidly. Noise then limits further extraction of information from  
these higher frequencies.

Hope this helps

Mark


On 15/04/2011, at 8:41 AM, Martin Wessendorf wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> On 4/14/2011 3:31 PM, Mark Cannell wrote:
>
>> I'm afraid the idea is MUCH older than 2008. I recall reading a paper
>> from the (?) 1960's that discussed the idea that the diffraction  
>> limit
>> was not a real limit at all. I believe they were using bacteria as an
>> example of where knowledge of the object imparts more information to
>> overcome the 'limit'
>
> I think no one would deny that if you know what you're looking at,  
> you can break the "resolution barrier"--STORM and PALM being prime  
> examples.  However, I'd still be interested in hearing whether--
> without having any additional information about the sample--there is  
> any theoretical limit to resolution in fluorescence microscopy,  
> other than that imposed by signal-to-noise ratio.
>
> Martin
> --
> Martin Wessendorf, Ph.D.                   office: [phone]
> Assoc Prof, Dept Neuroscience                 lab: [phone]
> University of Minnesota             Preferred FAX: [phone]
> 6-145 Jackson Hall, 321 Church St. SE    Dept Fax: [phone]
> Minneapolis, MN  55455                    e-mail: [hidden email]

[Steffen Dietzel]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

On 14.04.2011 22:38, Mark Cannell wrote:
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> It's multiphoton excitation so the effective PSF is much smaller than
> the 1 photon PSF. This has been described in numerous texts.

Mark, I am glad to hear that. I didn't come across those texts, though.
Could you give me some references? The text books I tried doesn't seem
to cover this, or not in a way that is intelligible without a physics
degree.


Not wanting
> to rain on your parade but for THG the resolution should be better than
> you measured.

Oh, I am not short of reasons why the measured PSF could be worse than
theoretically achievable, starting with Ri mismatch. But again, if you
could be more specific, that would be helpful.

Thanks,
Steffen

>
> Cheers
>
>
>
> On 15/04/2011, at 5:52 AM, Steffen Dietzel wrote:
>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Hi everybody,
>>
>> somewhat related to the ongoing discussion on resolution, I came
>> across a puzzle today concerning the resolution of multi-photon
>> microscopy.
>>
>> I measured the resolution of our third harmonic generation (THG)
>> microscope and surprisingly I came up with a full width half maximum
>> (FWHM) slightly better than theory allows. Seems the most likely
>> explanation is I applied the wrong theory. But which one is the
>> correct one?
>>
>> The experiment:
>> THG with 1275 nm, Objective 0.95 NA (water, 20x), beads 60 nm in 2%
>> agarose, voxel size 0.136 x 0.136 x 0.5 µm. Result: FWHM ~0.7 µm (for
>> both, forward and backward THG)
>>
>> Assuming that for multi-photon point-scanners only the excitation
>> wavelength is relevant, I used 1275 nm for the theory (Rayleigh):
>> r=0.61λ/NA = 0.82 µm
>>
>> So, the measured resolution is one pixel better than the theoretical
>> limit. You don't get that lucky every day ;-)
>>
>> Possibilities I have considered:
>> - I messed up the experiment. I wouldn't know, however, how I could
>> get a better result by messing up.
>> - Microscope settings are wrong (wrong pixel size). Possible of course
>> but not very likely.
>> - Rayleigh does not apply to multi-photon, I overlooked something. If
>> so, please help out.
>> - THG requires 3 photons to take place. Maybe the photon density is
>> low enough in the outer areas of the PSF so that signal generation is
>> limited to inner areas of the PSF? (Now that would be really
>> interesting from an academic point of view, since it would mean you
>> could squeeze the size of the excitation spot relative to the
>> wavelength with 4, 5, etc. photon effects. Although it probably
>> wouldn't do much good for practical purposes since you would have to
>> start with long wavelengths to end up with a visible (=easy
>> detectable) signal.)
>>
>>
>> Any ideas? Could people share measured FWHMs from their multi-photon
>> setup? Maybe even from a 3 photon process?
>>
>> Steffen
>>
>>
>> --
>> ------------------------------------------------------------
>> Steffen Dietzel, PD Dr. rer. nat
>> Ludwig-Maximilians-Universität München
>> Walter-Brendel-Zentrum für experimentelle Medizin (WBex)
>> Head of light microscopy
>>
>> Mail room:
>> Marchioninistr. 15, D-81377 München
>>
>> Building location:
>> Marchioninistr. 27, München-Großhadern
>

[Mark Cannell]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Stefan

Here's a starting point:

Practical Limits of Resolution in Confocal and
Non-Linear Microscopy
GUY COX AND COLIN J.R. SHEPPARD
MICROSCOPY RESEARCH AND TECHNIQUE 63:18 –22 (2004) (check out links  
from that paper too).

I think its discussed in the Handbook of biological confocal  
microscopy (Pawley) as well. The importance of RI mismatch cannot be  
overstated as the excitation volume is entirely dependent on the  
focussing ability of the objective.

Hope this helps, Mark


On 15/04/2011, at 7:25 PM, Steffen Dietzel wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> On 14.04.2011 22:38, Mark Cannell wrote:
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> It's multiphoton excitation so the effective PSF is much smaller than
>> the 1 photon PSF. This has been described in numerous texts.
>
> Mark, I am glad to hear that. I didn't come across those texts,  
> though. Could you give me some references? The text books I tried  
> doesn't seem to cover this, or not in a way that is intelligible  
> without a physics degree.
>
>
> Not wanting
>> to rain on your parade but for THG the resolution should be better  
>> than
>> you measured.
>
> Oh, I am not short of reasons why the measured PSF could be worse  
> than theoretically achievable, starting with Ri mismatch. But again,  
> if you could be more specific, that would be helpful.
>
> Thanks,
> Steffen
>
>
>>
>> Cheers
>>
>>
>>
>> On 15/04/2011, at 5:52 AM, Steffen Dietzel wrote:
>>
>>> *****
>>> To join, leave or search the confocal microscopy listserv, go to:
>>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>> *****
>>>
>>> Hi everybody,
>>>
>>> somewhat related to the ongoing discussion on resolution, I came
>>> across a puzzle today concerning the resolution of multi-photon
>>> microscopy.
>>>
>>> I measured the resolution of our third harmonic generation (THG)
>>> microscope and surprisingly I came up with a full width half maximum
>>> (FWHM) slightly better than theory allows. Seems the most likely
>>> explanation is I applied the wrong theory. But which one is the
>>> correct one?
>>>
>>> The experiment:
>>> THG with 1275 nm, Objective 0.95 NA (water, 20x), beads 60 nm in 2%
>>> agarose, voxel size 0.136 x 0.136 x 0.5 µm. Result: FWHM ~0.7 µm  
>>> (for
>>> both, forward and backward THG)
>>>
>>> Assuming that for multi-photon point-scanners only the excitation
>>> wavelength is relevant, I used 1275 nm for the theory (Rayleigh):
>>> r=0.61λ/NA = 0.82 µm
>>>
>>> So, the measured resolution is one pixel better than the theoretical
>>> limit. You don't get that lucky every day ;-)
>>>
>>> Possibilities I have considered:
>>> - I messed up the experiment. I wouldn't know, however, how I could
>>> get a better result by messing up.
>>> - Microscope settings are wrong (wrong pixel size). Possible of  
>>> course
>>> but not very likely.
>>> - Rayleigh does not apply to multi-photon, I overlooked something.  
>>> If
>>> so, please help out.
>>> - THG requires 3 photons to take place. Maybe the photon density is
>>> low enough in the outer areas of the PSF so that signal generation  
>>> is
>>> limited to inner areas of the PSF? (Now that would be really
>>> interesting from an academic point of view, since it would mean you
>>> could squeeze the size of the excitation spot relative to the
>>> wavelength with 4, 5, etc. photon effects. Although it probably
>>> wouldn't do much good for practical purposes since you would have to
>>> start with long wavelengths to end up with a visible (=easy
>>> detectable) signal.)
>>>
>>>
>>> Any ideas? Could people share measured FWHMs from their multi-photon
>>> setup? Maybe even from a 3 photon process?
>>>
>>> Steffen
>>>
>>>
>>> --
>>> ------------------------------------------------------------
>>> Steffen Dietzel, PD Dr. rer. nat
>>> Ludwig-Maximilians-Universität München
>>> Walter-Brendel-Zentrum für experimentelle Medizin (WBex)
>>> Head of light microscopy
>>>
>>> Mail room:
>>> Marchioninistr. 15, D-81377 München
>>>
>>> Building location:
>>> Marchioninistr. 27, München-Großhadern
>>

[Guy Cox-2]

Thanks for the plug, Mark!  What 2-photon does is give you the psf squared – so the resolution is improved by root 2.  The formula becomes 0.61 lambda / NA x 1.414.  In principle that’s what confocal does, but this requires an infinitely small pinhole and so cannot be achieved.  In 2P it is automatic.  What Colin and I did in that paper is calculate what resolution you could actually get with pinholes of different finite size.  We showed (experimentally and theoretically) that the root 2 resolution improvement makes 2-photon nearly as good as real-world confocal in spite of the longer wavelength.  Presumably in 3-photon you would get a root 3 (1.732) resolution improvement.  

 

                                                                      Guy

 

Sponsor my next half-marathon on May 15th

There’s a special reason – find it out at

http://www.everydayhero.com.au/Guy_Cox_4846 <http://www.everydayhero.com.au/Guy_Cox_4846>

______________________________________________

Associate Professor Guy Cox, MA, DPhil(Oxon)

Australian Centre for Microscopy & Microanalysis,

Madsen Building F09, University of Sydney, NSW 2006

 

Phone [phone]     Fax [phone]

             Mobile [phone]

______________________________________________

      http://www.guycox.net <http://www.guycox.net/>

 

From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Mark Cannell
Sent: Friday, 15 April 2011 8:22 PM
To: [hidden email]
Subject: Re: Resolution/PSF/FWHM in multi-photon microscopy

 

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Stefan

Here's a starting point:

Practical Limits of Resolution in Confocal and
Non-Linear Microscopy
GUY COX AND COLIN J.R. SHEPPARD
MICROSCOPY RESEARCH AND TECHNIQUE 63:18 –22 (2004) (check out links
from that paper too).

I think its discussed in the Handbook of biological confocal
microscopy (Pawley) as well. The importance of RI mismatch cannot be
overstated as the excitation volume is entirely dependent on the
focussing ability of the objective.

Hope this helps, Mark


On 15/04/2011, at 7:25 PM, Steffen Dietzel wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> On 14.04.2011 22:38, Mark Cannell wrote:
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> It's multiphoton excitation so the effective PSF is much smaller than
>> the 1 photon PSF. This has been described in numerous texts.
>
> Mark, I am glad to hear that. I didn't come across those texts,
> though. Could you give me some references? The text books I tried
> doesn't seem to cover this, or not in a way that is intelligible
> without a physics degree.
>
>
> Not wanting
>> to rain on your parade but for THG the resolution should be better
>> than
>> you measured.
>
> Oh, I am not short of reasons why the measured PSF could be worse
> than theoretically achievable, starting with Ri mismatch. But again,
> if you could be more specific, that would be helpful.
>
> Thanks,
> Steffen
>
>
>>
>> Cheers
>>
>>
>>
>> On 15/04/2011, at 5:52 AM, Steffen Dietzel wrote:
>>
>>> *****
>>> To join, leave or search the confocal microscopy listserv, go to:
>>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>> *****
>>>
>>> Hi everybody,
>>>
>>> somewhat related to the ongoing discussion on resolution, I came
>>> across a puzzle today concerning the resolution of multi-photon
>>> microscopy.
>>>
>>> I measured the resolution of our third harmonic generation (THG)
>>> microscope and surprisingly I came up with a full width half maximum
>>> (FWHM) slightly better than theory allows. Seems the most likely
>>> explanation is I applied the wrong theory. But which one is the
>>> correct one?
>>>
>>> The experiment:
>>> THG with 1275 nm, Objective 0.95 NA (water, 20x), beads 60 nm in 2%
>>> agarose, voxel size 0.136 x 0.136 x 0.5 µm. Result: FWHM ~0.7 µm
>>> (for
>>> both, forward and backward THG)
>>>
>>> Assuming that for multi-photon point-scanners only the excitation
>>> wavelength is relevant, I used 1275 nm for the theory (Rayleigh):
>>> r=0.61λ/NA = 0.82 µm
>>>
>>> So, the measured resolution is one pixel better than the theoretical
>>> limit. You don't get that lucky every day ;-)
>>>
>>> Possibilities I have considered:
>>> - I messed up the experiment. I wouldn't know, however, how I could
>>> get a better result by messing up.
>>> - Microscope settings are wrong (wrong pixel size). Possible of
>>> course
>>> but not very likely.
>>> - Rayleigh does not apply to multi-photon, I overlooked something.
>>> If
>>> so, please help out.
>>> - THG requires 3 photons to take place. Maybe the photon density is
>>> low enough in the outer areas of the PSF so that signal generation
>>> is
>>> limited to inner areas of the PSF? (Now that would be really
>>> interesting from an academic point of view, since it would mean you
>>> could squeeze the size of the excitation spot relative to the
>>> wavelength with 4, 5, etc. photon effects. Although it probably
>>> wouldn't do much good for practical purposes since you would have to
>>> start with long wavelengths to end up with a visible (=easy
>>> detectable) signal.)
>>>
>>>
>>> Any ideas? Could people share measured FWHMs from their multi-photon
>>> setup? Maybe even from a 3 photon process?
>>>
>>> Steffen
>>>
>>>
>>> --
>>> ------------------------------------------------------------
>>> Steffen Dietzel, PD Dr. rer. nat
>>> Ludwig-Maximilians-Universität München
>>> Walter-Brendel-Zentrum für experimentelle Medizin (WBex)
>>> Head of light microscopy
>>>
>>> Mail room:
>>> Marchioninistr. 15, D-81377 München
>>>
>>> Building location:
>>> Marchioninistr. 27, München-Großhadern
>>

________________________________

No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1209 / Virus Database: 1500/3574 - Release Date: 04/14/11

[Ian Dobbie]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Nuno Moreno <[hidden email]> writes:

> Therefore, I would be very, very skeptical either on any deconvolution
> and quantification using SIM.

Actually if your data has high S/N then quantitation should be
possible. For 2D super-resolution with SIM you have an over specified
linear problem, 9 images to give 4x the data (double resolution in
x&y). With 3D-SIM, as produced by OMX scopes, the problem is similar, 15
images producing 8x as much information (double res in x,y & z). Go look
at the original Gustafson papers for the details.

The issue is getting high enough S/N, which in practical terms means low
enough bleaching while taking those 100+ images.

Ian
--
Ian Dobbie
Micron Imaging Facility Manager,
Biochemistry,
University of Oxford,
South Parks Road,
Oxford
OX1 3QU
Tel: [phone]
Email: [hidden email]