Deconvolving Spinning Disk Images

[Dani Bodor]

Hi there,

I would like to deconvolve my spinning disk images for quantification purposes. Regarding this, I have a few issues/questions.

1. I have been trying to image fluorescent beads, in order to obtain PSF measurements for my microscope (conditions). However, whenever I do this, I always “see the grid” of the spinning disk unit, meaning that there are multiple points visible at the same time, even when a single bead is imaged. This only becomes obvious after severe rescaling and the other grid dots have a very weak intensity compared to the main dot (at least 2 orders of magnitude). I have tried synchronizing imaging time with spinning disk speed, long/short exposures, high/low laser intensities, with/without EM gain, etc. I always see the same thing.
The questions are: Will this affect the deconvolution efficiency/accuracy? Does anyone have any solution for this problem?



2. I have read conflicting opinions on performing deconvolution using experimental vs. theoretical PSFs vs maximum likelihood estimations (MLE). Some say that when imaging thick samples (in my case ~30um) it is better to do MLEs, because the PSF changes depending on the depth that you image, and measured PSFs are always on the surface of the slide. Others say that measured PSFs will always (usually) be better as they don’t assume anything but actually measure what is going on. Then again, I have been considering doing theoretical PSFs, as I have been having a lot of trouble actually measuring a proper PSF in my microscope (see point 1).
Does anyone have any experience with this or can advise me on these issues. Which programs/algorithms/plugins could I best use for this?

[Northan, Brian]

Commercial Response from Media Cybernetics:

I'm a software developer on the Autoquant product.

2.  The issue is not strictly theoretical vs. calculated but more of a question of how well you can measure or calculate the PSF.  For measured the issue is getting a picture of a bead under the same conditions (depth and RI) as the sample.  For theoretical it is knowing the values of the depth and sample RI.  Can you get a picture of a bead under conditions similar to the real experiment?? Or alternatively do you know a good estimate of depth and sample RI?? Both can be tricky.  

And actually MLE can be used with theoretical PSFs.  You may be thinking of the adaptive PSF MLE in Autoquant.  It's a version of MLE in which the PSF is allowed to change (subject to constraints).  Adaptive MLE uses a theoretical or measured PSF as a starting point.  Autoquant also has routines to detect spherical aberration (which can be problematic with thick samples).

Please feel free to contact us off list if you have any further questions.



-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Dani
Sent: Thursday, June 23, 2011 10:22 AM
To: [hidden email]
Subject: Deconvolving Spinning Disk Images

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

Hi there,

I would like to deconvolve my spinning disk images for quantification
purposes. Regarding this, I have a few issues/questions.

1. I have been trying to image fluorescent beads, in order to obtain PSF
measurements for my microscope (conditions). However, whenever I do this, I
always “see the grid” of the spinning disk unit, meaning that there are
multiple points visible at the same time, even when a single bead is imaged.
This only becomes obvious after severe rescaling and the other grid dots
have a very weak intensity compared to the main dot (at least 2 orders of
magnitude). I have tried synchronizing imaging time with spinning disk
speed, long/short exposures, high/low laser intensities, with/without EM
gain, etc. I always see the same thing.
The questions are: Will this affect the deconvolution efficiency/accuracy?
Does anyone have any solution for this problem?



2. I have read conflicting opinions on performing deconvolution using
experimental vs. theoretical PSFs vs maximum likelihood estimations (MLE).
Some say that when imaging thick samples (in my case ~30um) it is better to
do MLEs, because the PSF changes depending on the depth that you image, and
measured PSFs are always on the surface of the slide. Others say that
measured PSFs will always (usually) be better as they don’t assume anything
but actually measure what is going on. Then again, I have been considering
doing theoretical PSFs, as I have been having a lot of trouble actually
measuring a proper PSF in my microscope (see point 1).
Does anyone have any experience with this or can advise me on these issues.
Which programs/algorithms/plugins could I best use for this?


--
View this message in context: http://confocal-microscopy-list.588098.n2.nabble.com/Deconvolving-Spinning-Disk-Images-tp6508446p6508446.html
Sent from the Confocal Microscopy List mailing list archive at Nabble.com.

######################################################################################
CONFIDENTIALITY NOTICE:
This email transmission and its attachments contain confidential and proprietary information
of Princeton Instruments, Acton Research, Media Cybernetics and their affiliates and is
intended for the exclusive and confidential use of the intended recipient. Any use, dissemination,
printing, or copying of this transmission and its attachment(s) is strictly prohibited. If you
are not the intended recipient, please do not read, print, copy, distribute or take action in
reliance upon this message.  If you have received this in error, please notify the sender immediately
by telephone or return email and promptly delete all copies of the original transmission and its
attachments from your computer system.
#######################################################################################

[Gitta Hamel]

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

Commercial response from Huygens-SVI

Dear All,

Your 1st question:
The circular pattern is probably caused by bad synchronisation for which
you already did a number of smart tests. Normally such a pattern is
difficult to see in bead images so it looks like a real problem. I'll
email you off-line so that you can upload the data-set so that we can
help find the reason.

Your 2nd question:

I agree with Brian.  In the Huygens software, the MLE can be used both
with the measured and the (on the fly) theoretical PSF so the MLE
algorithm is actually not a deciding factor.
The deciding factors are indeed mainly the thickness and also the
possibly different Refractive Indices for lens and medium.

The rule of thumb is that up to 10 micron thickness  the measured PSF
can still be used. With thicker samples usage of the (on the fly)
theoretical PSF is usually better.  Still with identical RI's with
embedding media like Entallan or Aquatex you may go further than 10 micron.

So generally speaking a measured PSF is better than a theoretical PSF.
And as both spinning discs and confocal can suffer from bad alignments
that's an even stronger argument in favour of the measured one. In
Huygens you could apply for the first 10 micron the measured PSF and for
the remainder of your image the theoretical.

With best wishes,


Gitta Hamel

****************************************
Gitta Hamel
Managing Director Scientific Volume Imaging bv
Developers of the *HUYGENS* software
The Netherlands
phone: ++ 31 35 6 42 16 26
*****************************************


^SVI Customer support: mail us your questions [hidden email]
<mailto:[hidden email]>or find answers online in our Huygens
WIKI:www.svi.nl/FrontPage <http://%20www.svi.nl/FrontPage>

On 06/23/2011 05:18 PM, Northan, Brian wrote:

> Commercial Response from Media Cybernetics:
>
> I'm a software developer on the Autoquant product.
>
> 2.  The issue is not strictly theoretical vs. calculated but more of a question of how well you can measure or calculate the PSF.  For measured the issue is getting a picture of a bead under the same conditions (depth and RI) as the sample.  For theoretical it is knowing the values of the depth and sample RI.  Can you get a picture of a bead under conditions similar to the real experiment?? Or alternatively do you know a good estimate of depth and sample RI?? Both can be tricky.
>
> And actually MLE can be used with theoretical PSFs.  You may be thinking of the adaptive PSF MLE in Autoquant.  It's a version of MLE in which the PSF is allowed to change (subject to constraints).  Adaptive MLE uses a theoretical or measured PSF as a starting point.  Autoquant also has routines to detect spherical aberration (which can be problematic with thick samples).
>
> Please feel free to contact us off list if you have any further questions.
>
>
>
> -----Original Message-----
> From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Dani
> Sent: Thursday, June 23, 2011 10:22 AM
> To: [hidden email]
> Subject: Deconvolving Spinning Disk Images
>
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hi there,
>
> I would like to deconvolve my spinning disk images for quantification
> purposes. Regarding this, I have a few issues/questions.
>
> 1. I have been trying to image fluorescent beads, in order to obtain PSF
> measurements for my microscope (conditions). However, whenever I do this, I
> always “see the grid” of the spinning disk unit, meaning that there are
> multiple points visible at the same time, even when a single bead is imaged.
> This only becomes obvious after severe rescaling and the other grid dots
> have a very weak intensity compared to the main dot (at least 2 orders of
> magnitude). I have tried synchronizing imaging time with spinning disk
> speed, long/short exposures, high/low laser intensities, with/without EM
> gain, etc. I always see the same thing.
> The questions are: Will this affect the deconvolution efficiency/accuracy?
> Does anyone have any solution for this problem?
>
>
>
> 2. I have read conflicting opinions on performing deconvolution using
> experimental vs. theoretical PSFs vs maximum likelihood estimations (MLE).
> Some say that when imaging thick samples (in my case ~30um) it is better to
> do MLEs, because the PSF changes depending on the depth that you image, and
> measured PSFs are always on the surface of the slide. Others say that
> measured PSFs will always (usually) be better as they don’t assume anything
> but actually measure what is going on. Then again, I have been considering
> doing theoretical PSFs, as I have been having a lot of trouble actually
> measuring a proper PSF in my microscope (see point 1).
> Does anyone have any experience with this or can advise me on these issues.
> Which programs/algorithms/plugins could I best use for this?
>
>
> --
> View this message in context: http://confocal-microscopy-list.588098.n2.nabble.com/Deconvolving-Spinning-Disk-Images-tp6508446p6508446.html
> Sent from the Confocal Microscopy List mailing list archive at Nabble.com.
>
> ######################################################################################
> CONFIDENTIALITY NOTICE:
> This email transmission and its attachments contain confidential and proprietary information
> of Princeton Instruments, Acton Research, Media Cybernetics and their affiliates and is
> intended for the exclusive and confidential use of the intended recipient. Any use, dissemination,
> printing, or copying of this transmission and its attachment(s) is strictly prohibited. If you
> are not the intended recipient, please do not read, print, copy, distribute or take action in
> reliance upon this message.  If you have received this in error, please notify the sender immediately
> by telephone or return email and promptly delete all copies of the original transmission and its
> attachments from your computer system.
> #######################################################################################

[David Baddeley]

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

Just a brief comment on point 1 - if your microscope is producing ghosting
artifacts (being able to 'see the grid') then the PSF you measure from the beads
is the 'correct' PSF - deconvolving with a PSF measurement which includes ghosts
*should* remove the ghosts from the real image as well, provided they relative
position of the ghosts remains constant over the field of view (probably also
subject to numeric/ noise constraints). That said, having the ghosts in the
first place is not going to make your images any better and if there' a hardware
solution I'm sure it would be worth correcting.

If you want to get a handle on the whether measured or theoretical/blind PSFs
are best in your case, you can try embedding beads in a block of agar and/or
putting beads on either side of your biological specimen.

cheers,
David



----- Original Message ----
From: Dani <[hidden email]>
To: [hidden email]
Sent: Fri, 24 June, 2011 2:22:24 AM
Subject: Deconvolving Spinning Disk Images

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

Hi there,

I would like to deconvolve my spinning disk images for quantification
purposes. Regarding this, I have a few issues/questions.

1. I have been trying to image fluorescent beads, in order to obtain PSF
measurements for my microscope (conditions). However, whenever I do this, I
always “see the grid” of the spinning disk unit, meaning that there are
multiple points visible at the same time, even when a single bead is imaged.
This only becomes obvious after severe rescaling and the other grid dots
have a very weak intensity compared to the main dot (at least 2 orders of
magnitude). I have tried synchronizing imaging time with spinning disk
speed, long/short exposures, high/low laser intensities, with/without EM
gain, etc. I always see the same thing.
The questions are: Will this affect the deconvolution efficiency/accuracy?
Does anyone have any solution for this problem?



2. I have read conflicting opinions on performing deconvolution using
experimental vs. theoretical PSFs vs maximum likelihood estimations (MLE).
Some say that when imaging thick samples (in my case ~30um) it is better to
do MLEs, because the PSF changes depending on the depth that you image, and
measured PSFs are always on the surface of the slide. Others say that
measured PSFs will always (usually) be better as they don’t assume anything
but actually measure what is going on. Then again, I have been considering
doing theoretical PSFs, as I have been having a lot of trouble actually
measuring a proper PSF in my microscope (see point 1).
Does anyone have any experience with this or can advise me on these issues.
Which programs/algorithms/plugins could I best use for this?


--
View this message in context:
http://confocal-microscopy-list.588098.n2.nabble.com/Deconvolving-Spinning-Disk-Images-tp6508446p6508446.html

Sent from the Confocal Microscopy List mailing list archive at Nabble.com.

[Michal Opas]

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

Hello Dear Listers,

We have run into a conundrum regarding live, semi-long term and fixable
labelling of two population of mouse ES cells for aggregation/sorting
out studies.
We would like to fix label cells (two cell strains), let them aggregate,
fix the aggregates after 3-4 days and be able to make them out with either
am ancient Bio-Rad MRC 600 Krypton/Argon (488, 568 and 647 nm) or
a prehistorical Zeiss Photomicroscope (DAPI/Hoechst @ ~350 nm, FITC @
~390 nm and TRITC (~440 nm) ot Texas Red (~620 nm).

Invitrogen/Molecular Probes
(http://www.invitrogen.com/site/us/en/home/References/Molecular-Probes-The-Handbook/Fluorescent-Tracers-of-Cell-Morphology-and-Fluid-Flow/Membrane-Permeant-Reactive-Tracers.html#head2)
offer quite few cell tracking probes but figuring out what does what is
difficult (for me, at least).
They say that CellTracker Green CMFDA (5-chloromethylfluorescein
diacetate) labels cells that are both fluorescent and viable for at
least 24 hours after loading.
We need three-four days, though.
Orange-fluorescent CellTracker Orange CMRA  - ditto.

CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succinimidyl Ester
is highly recommended but nothing is said of its fixability.
It is said that CellTrace Far Red DDAO-SE is a fixable,
far-red–fluorescent tracer for long-term cell labelling but from the
literature given I am not sure of its longevity.

Sigma
(http://www.sigmaaldrich.com/img/assets/8200/PKH_LinkerKit_web_version.pdf)
sells PKH26 and PKH67 that (Sigma says) are stable for up to 100 days in
vivo (PKH26), but I do not see them heavily used in cell tracking other
than FACS.

/Ergo/,
Any practical suggestion shall be greatly appreciated!

Greetings,

Michal

PS
Obviously (and to my chagrin) I have no financial association with
neither company.
M.O.

[Dani Bodor]

> David: that is a point I had also already considered. Maybe it is in fact best to use the measured "ghosted" PSF for deconvolution. As I had already mentioned initially, the ghosts are much much less bright than the focal point, so in my images it will probably not be a huge problem. I am just afraid that small errors in the PSF measurements might heavily screw up the deconvolution process.

> Getta: thank you very much for your suggestions, i will try to find some samples pictures to send you so that you could take a look and hopefully advise me further.

[Lutz Schaefer]

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

Hello,

it seems to me that you are dealing with the common problem of pinhole cross
talk, in which the neighboring pinholes contribute in such way that your PSF
will contain next to the main peak about six more of lesser intensity around
it. To deconvolve such image, you need this patterned PSF! If you want to
use a theoretical PSF you need to work out the exact forward problem and
then generate a PSF that matches your instrument in all parameters
especially the geometry of the cross-talking pinholes. Working with blind
deconvolution might be tricky, as there are often heavy constraints on the
PSF to introduce additional knowledge. Unless your blind deconvolution
system supports such crosstalk PSF's, I would not resort to this option.
Finally you can try to acquire an empirical PSF using a small bead. The
caveat is, that you can only use a bead that is far enough away from others
so that it does not interfere with the crosstalk pattern of other beads.
Then, you just cut out and center this single bead stack and use it for
deconvolution.

hope it helped
regards
Lutz

__________________________________
L u t z   S c h a e f e r
Sen. Scientist
Mathematical modeling / Image processing
Advanced Imaging Methodology Consultation
16-715 Doon Village Rd.
Kitchener, ON, N2P 2A2, Canada
Phone/Fax: [phone]
Email:     [hidden email]
___________________________________

--------------------------------------------------
From: "Dani" <[hidden email]>
Sent: Thursday, June 23, 2011 10:22
To: <[hidden email]>
Subject: Deconvolving Spinning Disk Images

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hi there,
>
> I would like to deconvolve my spinning disk images for quantification
> purposes. Regarding this, I have a few issues/questions.
>
> 1. I have been trying to image fluorescent beads, in order to obtain PSF
> measurements for my microscope (conditions). However, whenever I do this,
> I
> always “see the grid” of the spinning disk unit, meaning that there are
> multiple points visible at the same time, even when a single bead is
> imaged.
> This only becomes obvious after severe rescaling and the other grid dots
> have a very weak intensity compared to the main dot (at least 2 orders of
> magnitude). I have tried synchronizing imaging time with spinning disk
> speed, long/short exposures, high/low laser intensities, with/without EM
> gain, etc. I always see the same thing.
> The questions are: Will this affect the deconvolution efficiency/accuracy?
> Does anyone have any solution for this problem?
>
>
>
> 2. I have read conflicting opinions on performing deconvolution using
> experimental vs. theoretical PSFs vs maximum likelihood estimations (MLE).
> Some say that when imaging thick samples (in my case ~30um) it is better
> to
> do MLEs, because the PSF changes depending on the depth that you image,
> and
> measured PSFs are always on the surface of the slide. Others say that
> measured PSFs will always (usually) be better as they don’t assume
> anything
> but actually measure what is going on. Then again, I have been considering
> doing theoretical PSFs, as I have been having a lot of trouble actually
> measuring a proper PSF in my microscope (see point 1).
> Does anyone have any experience with this or can advise me on these
> issues.
> Which programs/algorithms/plugins could I best use for this?
>
>
> --
> View this message in context:
> http://confocal-microscopy-list.588098.n2.nabble.com/Deconvolving-Spinning-Disk-Images-tp6508446p6508446.html
> Sent from the Confocal Microscopy List mailing list archive at Nabble.com.

[Kilgore, Jason-2]

** Vendor Response **

Hi, Michal,

The dyes that are SE (succinimidyl ester) derivatives will bind to proteins (via their lysine residues) in the cytoplasm, covalently, and thus are retained long-term.  3-4 days is not a problem with these.  CFDA SE is the most commonly used.  These are fixable due to that protein label and crosslinking with aldehydes.

Most of the CellTracker dyes also will bind to proteins as well, via thiol groups, and be retained over at least 4 days as well (with the exception of CellTracker BODIPY and CellTracker CM-DiI, which are lipophilic membrane stains).  They are also fixable, as with the SE dyes.

I would recommend avoiding the lipophilic cyanine dyes, including the PKH kits, due to the fact that they just intercalate into the membranes without covalently binding to cell constituents, and are not as fixable.

Another choice is to use Qdots, namely our Qtracker cell labeling reagents, which are actively taken up by endocytosis and sequestered into endosomes.  These, too, are retained with fixation (and can even survive paraffin processing).

It looks like you have filter options for a number of these choices.

If you would like more help in choosing, please let me know, offline.

Cheers,

Jason


Jason A. Kilgore
Technical Application Scientist
Molecular Probes Labeling and Detection Technologies
Cells Systems Division
 
T 1 800 955 6288 then option 5  or  541 335 0353 • F 541 335 0238
29851 Willow Creek Rd • Eugene • OR • 97402-9132 • United States
www.invitrogen.com/technicalsupport

 



-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Michal Opas
Sent: Thursday, June 23, 2011 2:39 PM
To: [hidden email]
Subject: Live fixable fluorescent cell labelling

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

Hello Dear Listers,

We have run into a conundrum regarding live, semi-long term and fixable
labelling of two population of mouse ES cells for aggregation/sorting
out studies.
We would like to fix label cells (two cell strains), let them aggregate,
fix the aggregates after 3-4 days and be able to make them out with either
am ancient Bio-Rad MRC 600 Krypton/Argon (488, 568 and 647 nm) or
a prehistorical Zeiss Photomicroscope (DAPI/Hoechst @ ~350 nm, FITC @
~390 nm and TRITC (~440 nm) ot Texas Red (~620 nm).

Invitrogen/Molecular Probes
(http://www.invitrogen.com/site/us/en/home/References/Molecular-Probes-The-Handbook/Fluorescent-Tracers-of-Cell-Morphology-and-Fluid-Flow/Membrane-Permeant-Reactive-Tracers.html#head2)
offer quite few cell tracking probes but figuring out what does what is
difficult (for me, at least).
They say that CellTracker Green CMFDA (5-chloromethylfluorescein
diacetate) labels cells that are both fluorescent and viable for at
least 24 hours after loading.
We need three-four days, though.
Orange-fluorescent CellTracker Orange CMRA  - ditto.

CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succinimidyl Ester
is highly recommended but nothing is said of its fixability.
It is said that CellTrace Far Red DDAO-SE is a fixable,
far-red–fluorescent tracer for long-term cell labelling but from the
literature given I am not sure of its longevity.

Sigma
(http://www.sigmaaldrich.com/img/assets/8200/PKH_LinkerKit_web_version.pdf)
sells PKH26 and PKH67 that (Sigma says) are stable for up to 100 days in
vivo (PKH26), but I do not see them heavily used in cell tracking other
than FACS.

/Ergo/,
Any practical suggestion shall be greatly appreciated!

Greetings,

Michal

PS
Obviously (and to my chagrin) I have no financial association with
neither company.
M.O.

[James Pawley]

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

Well, what Lutz says is precisely true, but on
the other hand, unless the stained part of the
specimen is quite thick, it may not be very
important.

Yes, light that originates far from the focus
plane may come through the wrong pinhole. As the
pinhole in an image plane it will actually arrive
at the CCD in the wrong location. But on the
other hand, it isn't very much light. In the
Yoko, where the pinholes are 50µm and 250µm
apart, the circle of transmission where the first
light would come through has a circumference
about 1570 µm (of which at most 6x50 or 300 µmm
is open, more likely about 200 µm),  a diameter
of 500 µm and an area of about 200k square
microns. Of this, about 8000 square microns is
"hole", about 4%. Apart from the geometry, the
light focused back from the "spot" in the
specimen tends to cluster near the pinhole, so
the fraction of this out-of-focus light
prohibited from reaching the detector is probably
more than 96 parts out of 100.

It is then important to remember that this is a
confocal microscope, and we often only get very
few photons through the pinhole in any case
(maybe tens-hundreds  in a 1 sec exposure) and "a
lot less that very few is close to zero" if you
will permit the imprecision.

The z-resolution of disc scanners was much
studied when they first came out and the Z-image
of a plane of fluorophor is almost the same as a
single point scanner(with the same pinhole size)
with the exception that when the focus plane
moves far from focus, the signal level does not
asymptote to zero, but to the amount of light
that would pass an ND filter with the same
average transmission as the disk. Other stain
distributions may be produce other results (a
periodic structure that appeared in the plane of
the disk with a period that matched that of the
hole pattern for instance) but it is not beyond
the wit of man to obtain and use a measured 3D
PSF to deconvolve confocal data of any type,
including disc scanners.

I recommend it.

JP

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Hello,
>
>it seems to me that you are dealing with the
>common problem of pinhole cross talk, in which
>the neighboring pinholes contribute in such way
>that your PSF will contain next to the main peak
>about six more of lesser intensity around it. To
>deconvolve such image, you need this patterned
>PSF! If you want to use a theoretical PSF you
>need to work out the exact forward problem and
>then generate a PSF that matches your instrument
>in all parameters especially the geometry of the
>cross-talking pinholes. Working with blind
>deconvolution might be tricky, as there are
>often heavy constraints on the PSF to introduce
>additional knowledge. Unless your blind
>deconvolution system supports such crosstalk
>PSF's, I would not resort to this option.
>Finally you can try to acquire an empirical PSF
>using a small bead. The caveat is, that you can
>only use a bead that is far enough away from
>others so that it does not interfere with the
>crosstalk pattern of other beads. Then, you just
>cut out and center this single bead stack and
>use it for deconvolution.
>
>hope it helped
>regards
>Lutz
>
>__________________________________
>L u t z   S c h a e f e r
>Sen. Scientist
>Mathematical modeling / Image processing
>Advanced Imaging Methodology Consultation
>16-715 Doon Village Rd.
>Kitchener, ON, N2P 2A2, Canada
>Phone/Fax: [phone]
>Email:     [hidden email]
>___________________________________
>
>--------------------------------------------------
>From: "Dani" <[hidden email]>
>Sent: Thursday, June 23, 2011 10:22
>To: <[hidden email]>
>Subject: Deconvolving Spinning Disk Images
>
>>*****
>>To join, leave or search the confocal microscopy listserv, go to:
>>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>*****
>>
>>Hi there,
>>
>>I would like to deconvolve my spinning disk images for quantification
>>purposes. Regarding this, I have a few issues/questions.
>>
>>1. I have been trying to image fluorescent beads, in order to obtain PSF
>>measurements for my microscope (conditions). However, whenever I do this, I
>>always "see the grid" of the spinning disk unit, meaning that there are
>>multiple points visible at the same time, even when a single bead is imaged.
>>This only becomes obvious after severe rescaling and the other grid dots
>>have a very weak intensity compared to the main dot (at least 2 orders of
>>magnitude). I have tried synchronizing imaging time with spinning disk
>>speed, long/short exposures, high/low laser intensities, with/without EM
>>gain, etc. I always see the same thing.
>>The questions are: Will this affect the deconvolution efficiency/accuracy?
>>Does anyone have any solution for this problem?
>>
>>
>>
>>2. I have read conflicting opinions on performing deconvolution using
>>experimental vs. theoretical PSFs vs maximum likelihood estimations (MLE).
>>Some say that when imaging thick samples (in my case ~30um) it is better to
>>do MLEs, because the PSF changes depending on the depth that you image, and
>>measured PSFs are always on the surface of the slide. Others say that
>>measured PSFs will always (usually) be better as they don't assume anything
>>but actually measure what is going on. Then again, I have been considering
>>doing theoretical PSFs, as I have been having a lot of trouble actually
>>measuring a proper PSF in my microscope (see point 1).
>>Does anyone have any experience with this or can advise me on these issues.
>>Which programs/algorithms/plugins could I best use for this?
>>
>>
>>--
>>View this message in context:
>>http://confocal-microscopy-list.588098.n2.nabble.com/Deconvolving-Spinning-Disk-Images-tp6508446p6508446.html
>>Sent from the Confocal Microscopy List mailing list archive at Nabble.com.

--
                    ****************************************
Prof. James B. Pawley,                                     Phone: [phone]
3D Microscopy of Living Cells: Summer Course   CELL: 778-919-3176
Info at: http://www.3dcourse.ubc.ca
"If it isn't diffraction, it is statistics":Microscopist's complaint, Anon.

[Paul Maddox]

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

I have posted this before, but will repost.  We wrote a methods chapter on
spinning disk in 2003 (Maddox et al., methods in Enzymology) where we
discussed several of these issues.  I would be happy to provide a PDF for
those interested if you email me offline  ([hidden email]).

Happy imaging!

Paul


On 6/23/11 8:31 PM, "James Pawley" <[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Well, what Lutz says is precisely true, but on
> the other hand, unless the stained part of the
> specimen is quite thick, it may not be very
> important.
>
> Yes, light that originates far from the focus
> plane may come through the wrong pinhole. As the
> pinhole in an image plane it will actually arrive
> at the CCD in the wrong location. But on the
> other hand, it isn't very much light. In the
> Yoko, where the pinholes are 50µm and 250µm
> apart, the circle of transmission where the first
> light would come through has a circumference
> about 1570 µm (of which at most 6x50 or 300 µmm
> is open, more likely about 200 µm),  a diameter
> of 500 µm and an area of about 200k square
> microns. Of this, about 8000 square microns is
> "hole", about 4%. Apart from the geometry, the
> light focused back from the "spot" in the
> specimen tends to cluster near the pinhole, so
> the fraction of this out-of-focus light
> prohibited from reaching the detector is probably
> more than 96 parts out of 100.
>
> It is then important to remember that this is a
> confocal microscope, and we often only get very
> few photons through the pinhole in any case
> (maybe tens-hundreds  in a 1 sec exposure) and "a
> lot less that very few is close to zero" if you
> will permit the imprecision.
>
> The z-resolution of disc scanners was much
> studied when they first came out and the Z-image
> of a plane of fluorophor is almost the same as a
> single point scanner(with the same pinhole size)
> with the exception that when the focus plane
> moves far from focus, the signal level does not
> asymptote to zero, but to the amount of light
> that would pass an ND filter with the same
> average transmission as the disk. Other stain
> distributions may be produce other results (a
> periodic structure that appeared in the plane of
> the disk with a period that matched that of the
> hole pattern for instance) but it is not beyond
> the wit of man to obtain and use a measured 3D
> PSF to deconvolve confocal data of any type,
> including disc scanners.
>
> I recommend it.
>
> JP
>
>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Hello,
>>
>> it seems to me that you are dealing with the
>> common problem of pinhole cross talk, in which
>> the neighboring pinholes contribute in such way
>> that your PSF will contain next to the main peak
>> about six more of lesser intensity around it. To
>> deconvolve such image, you need this patterned
>> PSF! If you want to use a theoretical PSF you
>> need to work out the exact forward problem and
>> then generate a PSF that matches your instrument
>> in all parameters especially the geometry of the
>> cross-talking pinholes. Working with blind
>> deconvolution might be tricky, as there are
>> often heavy constraints on the PSF to introduce
>> additional knowledge. Unless your blind
>> deconvolution system supports such crosstalk
>> PSF's, I would not resort to this option.
>> Finally you can try to acquire an empirical PSF
>> using a small bead. The caveat is, that you can
>> only use a bead that is far enough away from
>> others so that it does not interfere with the
>> crosstalk pattern of other beads. Then, you just
>> cut out and center this single bead stack and
>> use it for deconvolution.
>>
>> hope it helped
>> regards
>> Lutz
>>
>> __________________________________
>> L u t z   S c h a e f e r
>> Sen. Scientist
>> Mathematical modeling / Image processing
>> Advanced Imaging Methodology Consultation
>> 16-715 Doon Village Rd.
>> Kitchener, ON, N2P 2A2, Canada
>> Phone/Fax: [phone]
>> Email:     [hidden email]
>> ___________________________________
>>
>> --------------------------------------------------
>> From: "Dani" <[hidden email]>
>> Sent: Thursday, June 23, 2011 10:22
>> To: <[hidden email]>
>> Subject: Deconvolving Spinning Disk Images
>>
>>> *****
>>> To join, leave or search the confocal microscopy listserv, go to:
>>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>> *****
>>>
>>> Hi there,
>>>
>>> I would like to deconvolve my spinning disk images for quantification
>>> purposes. Regarding this, I have a few issues/questions.
>>>
>>> 1. I have been trying to image fluorescent beads, in order to obtain PSF
>>> measurements for my microscope (conditions). However, whenever I do this, I
>>> always "see the grid" of the spinning disk unit, meaning that there are
>>> multiple points visible at the same time, even when a single bead is imaged.
>>> This only becomes obvious after severe rescaling and the other grid dots
>>> have a very weak intensity compared to the main dot (at least 2 orders of
>>> magnitude). I have tried synchronizing imaging time with spinning disk
>>> speed, long/short exposures, high/low laser intensities, with/without EM
>>> gain, etc. I always see the same thing.
>>> The questions are: Will this affect the deconvolution efficiency/accuracy?
>>> Does anyone have any solution for this problem?
>>>
>>>
>>>
>>> 2. I have read conflicting opinions on performing deconvolution using
>>> experimental vs. theoretical PSFs vs maximum likelihood estimations (MLE).
>>> Some say that when imaging thick samples (in my case ~30um) it is better to
>>> do MLEs, because the PSF changes depending on the depth that you image, and
>>> measured PSFs are always on the surface of the slide. Others say that
>>> measured PSFs will always (usually) be better as they don't assume anything
>>> but actually measure what is going on. Then again, I have been considering
>>> doing theoretical PSFs, as I have been having a lot of trouble actually
>>> measuring a proper PSF in my microscope (see point 1).
>>> Does anyone have any experience with this or can advise me on these issues.
>>> Which programs/algorithms/plugins could I best use for this?
>>>
>>>
>>> --
>>> View this message in context:
>>> http://confocal-microscopy-list.588098.n2.nabble.com/Deconvolving-Spinning-D
>>> isk-Images-tp6508446p6508446.html
>>> Sent from the Confocal Microscopy List mailing list archive at Nabble.com.
>
>
> --
>                     ****************************************
> Prof. James B. Pawley,                                     Phone: [phone]
> 3D Microscopy of Living Cells: Summer Course   CELL: 778-919-3176
> Info at: http://www.3dcourse.ubc.ca
> "If it isn't diffraction, it is statistics":Microscopist's complaint, Anon.

[JOEL B. SHEFFIELD]

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

In the dark ages...I published a study some time ago in which I used
rhodamine or fluorescein isothiocyanate to label separate populations of
cells, whose behavior in aggregates was then monitored by fluorescence.
Here's the reference:  Developmental Biology 89: 41-47 (1982)

On Thu, Jun 23, 2011 at 5:38 PM, Michal Opas <[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/**wa?A0=confocalmicroscopy<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
> *****
>
> Hello Dear Listers,
>
> We have run into a conundrum regarding live, semi-long term and fixable
> labelling of two population of mouse ES cells for aggregation/sorting out
> studies.
> We would like to fix label cells (two cell strains), let them aggregate,
> fix the aggregates after 3-4 days and be able to make them out with either
> am ancient Bio-Rad MRC 600 Krypton/Argon (488, 568 and 647 nm) or
> a prehistorical Zeiss Photomicroscope (DAPI/Hoechst @ ~350 nm, FITC @ ~390
> nm and TRITC (~440 nm) ot Texas Red (~620 nm).
>
> Invitrogen/Molecular Probes (http://www.invitrogen.com/**
> site/us/en/home/References/**Molecular-Probes-The-Handbook/**
> Fluorescent-Tracers-of-Cell-**Morphology-and-Fluid-Flow/**
> Membrane-Permeant-Reactive-**Tracers.html#head2<http://www.invitrogen.com/site/us/en/home/References/Molecular-Probes-The-Handbook/Fluorescent-Tracers-of-Cell-Morphology-and-Fluid-Flow/Membrane-Permeant-Reactive-Tracers.html#head2>)
> offer quite few cell tracking probes but figuring out what does what is
> difficult (for me, at least).
> They say that CellTracker Green CMFDA (5-chloromethylfluorescein diacetate)
> labels cells that are both fluorescent and viable for at least 24 hours
> after loading.
> We need three-four days, though.
> Orange-fluorescent CellTracker Orange CMRA  - ditto.
>
> CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succinimidyl Ester is
> highly recommended but nothing is said of its fixability.
> It is said that CellTrace Far Red DDAO-SE is a fixable, far-red–fluorescent
> tracer for long-term cell labelling but from the literature given I am not
> sure of its longevity.
>
> Sigma (http://www.sigmaaldrich.com/**img/assets/8200/PKH_LinkerKit_**
> web_version.pdf<http://www.sigmaaldrich.com/img/assets/8200/PKH_LinkerKit_web_version.pdf>)
> sells PKH26 and PKH67 that (Sigma says) are stable for up to 100 days in
> vivo (PKH26), but I do not see them heavily used in cell tracking other than
> FACS.
>
> /Ergo/,
> Any practical suggestion shall be greatly appreciated!
>
> Greetings,
>
> Michal
>
> PS
> Obviously (and to my chagrin) I have no financial association with neither
> company.
> M.O.
>

--


Joel B. Sheffield, Ph.D
Department of Biology
Temple University
Philadelphia, PA 19122
Voice: 215 204 8839
e-mail: [hidden email]
URL:  http://astro.temple.edu/~jbs

[Michal Opas]

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

Right!
Thanks Joel,

Michal

[JOEL B. SHEFFIELD]

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

I just looked over the paper, and I see that we were  able to retain label
only for two days or so because, we assumed, of protein turnover in the
cells.  This may be a fundamental problem that you will have to consider.
You may want to consider a fluorescent protein label that could be
transfected in.

Joel


On Fri, Jun 24, 2011 at 12:44 PM, Michal Opas <[hidden email]> wrote:

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

--


Joel B. Sheffield, Ph.D
Department of Biology
Temple University
Philadelphia, PA 19122
Voice: 215 204 8839
e-mail: [hidden email]
URL:  http://astro.temple.edu/~jbs

[Kilgore, Jason-2]

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

Likely TRITC and FITC would be surface labels, which might have higher protein turnover.  Another drawback of that method, for some assays, would be the potential disruption of cell surface protein function after the conjugation, potentially affecting cell-cell interactions.


Jason

Jason A. Kilgore
Technical Application Scientist
Molecular Probes Labeling and Detection Technologies
Cells Systems Division
 
T 1 800 955 6288 then option 5  or  541 335 0353 . F 541 335 0238
29851 Willow Creek Rd . Eugene . OR . 97402-9132 . United States
www.invitrogen.com/technicalsupport

 



-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of JOEL B. SHEFFIELD
Sent: Friday, June 24, 2011 9:50 AM
To: [hidden email]
Subject: Re: Live fixable fluorescent cell labelling

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

I just looked over the paper, and I see that we were  able to retain label
only for two days or so because, we assumed, of protein turnover in the
cells.  This may be a fundamental problem that you will have to consider.
You may want to consider a fluorescent protein label that could be
transfected in.

Joel


On Fri, Jun 24, 2011 at 12:44 PM, Michal Opas <[hidden email]> wrote:

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

--


Joel B. Sheffield, Ph.D
Department of Biology
Temple University
Philadelphia, PA 19122
Voice: 215 204 8839
e-mail: [hidden email]
URL:  http://astro.temple.edu/~jbs