Posted by
phil laissue-2 on
URL: http://confocal-microscopy-list.275.s1.nabble.com/Re-Assessing-phototoxicity-in-live-fluorescence-imaging-tp7587005p7587021.html
*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopyPost images on
http://www.imgur.com and include the link in your posting.
*****
Dear imagers,
Claire has raised an important issue, and it's great to see the discussion
and solutions - and so many people interested in reporting power
measurements. Thanks to everyone who chipped in with great advice.
There almost seem to be as many ways to do this as there are microscopists,
so this is a tricky area. We have used references in our review we thought
are reasonably practical:
- Waters, J.C. & Wittmann, T. (eds.) Quantitative Imaging in Cell Biology.
(Academic Press, 2014).
- Cranfill, P.J. et al. Nat. Methods 13, 557–562 (2016).
As discussed in this thread, conventional power sensors, where the flat
entry window is placed against the incoming light, are useful for measuring
the power of low NA objectives. They cannot be used to accurately measure
high-NA oil and water immersion lenses, which produce a highly focussed
cone of light (Dobrucki, 2013 - see reference below). The Thorlabs S170C
slide power sensor was developed with this problem in mind, with an
index-matching layer between the protective glass window and the photodiode
to prevent reflection, and can be used for high NA oil and water objectives.
- Fluorescence microscopy. JW Dobrucki. In: Fluorescence Microscopy: From
Principles to Biological Applications. First Edition. Edited by Ulrich
Kubitscheck. Wiley-VCH Verlag GmbH & Co. KGaA, 2013.
A practical approach using an iris, as mentioned by Zdenek, is described
here:
Grünwald D, Shenoy SM, Burke S, Singer RH. Calibrating excitation light
fluxes for quantitative light microscopy in cell biology. Nat
Protoc. 2008;3(11):1809-14. doi: 10.1038/nprot.2008.180. PubMed PMID:
18974739;
Note that there were similar threads in this forum a few years ago:
instrument to measure laser intensity on slide (05/07/2014)
Reporting laser power in publication (18/08/2015)
There is also a practical note by Vojnovic, Newman and Barber from 2007
(updated in 2011):
http://users.ox.ac.uk/~atdgroup/technicalnotes/Optical%20power%20meters%20for%20fluorescence%20microscopy.pdf
@ Richard Cole: I don't know the publication you mention, but it would be
great if you could post the reference here. Thanks!
Because of these differences and the technical aspects that need to be
considered, it is difficult for a routine lab to do this, and e.g. the
Thorlabs power sensor comes with a noticeable price tag. For these reasons,
it would be really *most *helpful if technology developers and commercial
microscopy manufacturers would provide data on the amounts of light
entering the sample at any given settings of the microscope, or to
incorporate tools to easily obtain such measurements, such that
non-specialists can do it without too much trouble. It's a shame that even
light transmission data for an objective should often be so hard to obtain.
So in many cases, a power measurement is clearly more of an estimate. It
might be useful for us as a community to decide on one standardised
approach. Until then, careful reporting on how it was done (while avoiding
the most prominent errors as described in this thread) will be the best way
forward.
With the best wishes,
Philippe
_________________________________________
Philippe Laissue, PhD
Royal Society Industry Fellow and MBL Whitman Center Scientist
University of Essex, Colchester CO4 3SQ, UK
(0044) 01206 872246 / (0044) 07842 676 456
[hidden email]
website <
https://laissue.github.io/>
On 19 July 2017 at 22:48, George McNamara <
[hidden email]> wrote:
reflection mode (low laser power!). If ambitious, measure with different
pinhole sizes. Zeiss has an RT80/20 (which I hope means 20% reflection, 80%
transmission), Olympus FV3000 has 10/90 (reflection/transmission).
>
> 2. Widefield microscopes, excitation side,
http://www.epitechnology.commanufactures (3D prints) filter cubes with a mirror in place of the
dichroic, and facing the 'other way', to enable imaging the lamp (or LLG or
fiber) onto the detector -- which you (or someone else) paid (a lot of)
money for, and should be quantitative. Examples of what you'll see are
shown on
http://www.epitechnology.com/epitechnology-1/>
> enjoy,
>
> George
>
>
> On 7/19/2017 4:08 PM, JAMES B PAWLEY wrote:
>>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>> Post images on
http://www.imgur.com and include the link in your posting.
>> *****
>>
>> Even the simplest things aren’t.
>>
>> Steffen is quite right about the power meter assuming (almost) normal
incidence. You might think you could reduce this problem by putting
immersion oil between the high-NA objective and the sensor surface,
however, in many power meters, the silicon detector is separated from the
protective cover glass by an air gap, so this doesn’t work (although you
might be able to get around this by using a home-made
>> power meter consisting of a small “oilable” silicon solar cell from the
electronics store connected to a current meter. But you would have to make
sure that no stray room light hits this usually rather large cell and you
would also have to calibrate it against a “real” power meter while
remembering to correct for their different areas…)
>>
>> The official way to do this measurement is to set up the microscope
optics for Kohler Illumination using an oiled condenser with an NA at least
as high as that of the objective. Assuming that there is no specimen in the
way, the light emerging from the condenser should be parallel to the axis.
Of course, you will still have reflection and absorption losses in the
condenser and if you are doing this in wide field, the setting of the field
diaphragm will have a massive effect (the transmitted power will be
proportional to the area of the image plane illuminated by the field
diaphragm). You will also have to be sure that the ray bundle from the
condenser isn’t larger than your sensor.
>>
>> On the other hand, as few scopes are set up with the required high-NA
oiled condenser, it is common to fall back on the “10X low-NA” option.
>>
>> Although this can provide fairly consistent day-to-day readings, you
should NOT assume that you can just apply this reading to what would have
emerged from a high-NA objective of higher magnification. This is because
the illumination system is set up to provide a beam of a certain size in
the BFP of the objective (in the space between the back of an
infinity-corrected objective and the tube lens). This beam is usually
roughly Gaussian in shape and its size is a compromise between the need to
“fill the NA” of the “best lens” with fair uniformity and the desire not
to waste too much light hitting the metal mounting of the optics (and then
scattering all over the place causing "stray light”).
>>
>> The problem is that “filling the NA” of, say a 40x 1.3 objective
requires a beam diameter that is 2.5x larger (and 6.25 larger in area) than
would be required for a 100x 1.3 objective. This is why early confocals
often got better resolution with higher mag objectives:The ray bundle just
didn’t fill the BFP of the lower-mag objectives. Conversely, they often got
better “penetration” when using oil lenses into aqueous specimens when
using the low-mag objectives: spherical aberration increases rapidly with
NA and by being underfilled, the lower mag objectives were actually
operating at a lower NA (at least on the illumination side).
>>
>> So, although you can use the "10x low-NA" trick for monitoring
day-to-day performance, you will probably need to use the Kohler set-up at
least once to determine how the illumination optics in your scope fills the
BFP. (It won’t work just to assume that the 40x 1.3 should pass 6.25x more
light than the 100x 1.3 because, even if the beam is large at the BFP, it
is still a Gaussian and brighter in the centre.) Basically, you need to
determine the real conversion factors between your low-NA 10x and all your
other objectives.
>>
>> Even using a “dry" NA 0.9 condenser is better than trying to measure the
convergent/divergent spot from a high NA objective. Try it with something
like a 1.3 NA 40x oil (or a 0.5NA 10x dry) and, after you have set up
Kohler using the normal transmitted light optics, open the condenser
aperture to 0.9, turn off the transmitted source and turn on the
epi-illumination (laser or WF) and hold a piece of white paper in the beam
coming out of the condenser. You should see a bright blob and this blob may
not even extend as far as NA 0.9 (which you can determine by moving the
condenser aperture control to see if the edges of the blob are cut off by a
sharp, black edge.) Knowing the size of this blob can help you at least
estimate how much of the other objectives will be "filled."
>>
>> I hope that this isn’t all too complicated and therefore disappointing
because I think that, particularly when working with living cells, nothing
is more important than knowing how much light hit the specimen while you
collected your images.
>>
>> So please persist!
>>
>> Jm Pawley
>>
>> James and Christine Pawley, 5446 Burley Place, Box 2348, Sechelt BC,
Canada, V0N3A0
[hidden email]<mailto:
[hidden email]>, Phone
1-604-885-0840 <(604)%20885-0840>, cell 1-604-989-6146 <(604)%20989-6146>
>>
>> James and Christine Pawley, 5446 Burley Place, Box 2348, Sechelt BC,
Canada, V0N3A0
[hidden email]<mailto:
[hidden email]>, Phone
1-604-885-0840 <(604)%20885-0840>, cell 1-604-989-6146 <(604)%20989-6146>
>>
>>
>>
>> On Jul 19, 2017, at 6:55 AM,
[hidden email]<mailto:
[hidden email]>
wrote:
>>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>> Post images on
http://www.imgur.com and include the link in your posting.
>> *****
>>
>> Hi Steffen,
>> you can improve the accuracy of the method "a)" (that is measuring the
laser
>> power before reaching the objective) greatly by mounting an iris stop in
>> place of the lens, adjusting the aperture to equal the diameter of the
back
>> focal plane (BFP) aperture of the objective lens, and measuring the
power of
>> the light that gets through this aperture.
>>
>> Most confocal microscopes 'overfill' the BFP greatly (which is good for
>> resolution) and you can get an order of magnitude difference when the
laser
>> beam is > 15 mm 'diameter' (it's gaussian profile) and the BFP diameter
is
>> just 4 mm (e.g. some high-magnification lenses).
>>
>> You can determine the BFP diameter by looking at the back of the lens,
or by
>> dong some simple math (I guess diameter = 2 * NA * focal_length; focal_
>> length = tube_lens_focal_lenght / magnification).
>>
>> This way, your a) and b) results should be much closer to each other and
to
>> the real value in between...
>>
>> Best, zdenek
>>
>> --
>> Zdenek Svindrych, Ph.D.
>> W.M. Keck Center for Cellular Imaging (PLSB 003)
>> Department of Biology,University of Virginia
>> 409 McCormick Rd, Charlottesville, VA-22904
>>
http://www.kcci.virginia.edu/>> tel: 434-982-4869 <(434)%20982-4869>
>>
>> ---------- Původní e-mail ----------
>> Od: Steffen Dietzel <
[hidden email]>
>> Komu:
[hidden email]
>> Datum: 19. 7. 2017 7:03:20
>> Předmět: Re: Assessing phototoxicity in live fluorescence imaging
>> "*****
>> To join, leave or search the confocal microscopy listserv, go to:
>>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>> Post images on
http://www.imgur.com and include the link in your posting.
>> *****
>>
>> Unfortunately not a solution, but a further complication: As far as I
>> know, power meters are made to detect light that hits the sensor
>> orthogonally. So with a high NA lens, a lot of the incident light won't
>> be even detected.
>>
>> I guess what one could do is to measure
>>
>> a) the power without objective with a parked beam, focusing on a spot in
>> the center of the field of view. This would give the upper estimate but
>> not the true intensity since some of it is absorbed by the objective
>> itself. Transmittance is never 100%.
>>
>> b) doing the same with the objective that is to be used. This will give
>> the lower estimate. Too low, since part of the light won't be measured
>> due to the incident angle.
>>
>> The truth then is somewhere between the two values. With modern high NA
>> objectives which should have a high transmission my gut feeling is that
>> the truth would be closer to (a) than to (b).
>>
>> You could take value (a) and correct it the transmission of the
>> objective at the given wavelength published by the manufacturer, if that
>> is available. But I don't think I have ever seen a paper that actually
>> did all that. Whatever value you take, as Andreas suggested you then
>> would have to relate it to the true pixel dwell time, i.e. disregarding
>> dead time of the scanner.
>>
>> To get the exact value at the focal point in the sample also would
>> require to take the losses due to reflection at the coverslip into
>> account. In essence, I am definitely with Claire when she says:
>>
>> It is my
>> thought that comparing relative powers on the same instrument is okay but
>> comparing between systems will be very complex.
>>
>> The Leica SP8 systems do allow to park the beam, as Craig suspected.
>> Since I always forget how to do that I put the procedure on our web
>> site, where I can easily find it :-)
>>
http://www.bioimaging.bmc.med.uni-muenchen.de/manuals-protocols/maintenance_
>> protocolls/laserpower/index.html
>>
>>
>> Cheers
>>
>> Steffen
>>
>>
>> Am 19.07.2017 um 00:24 schrieb Craig Brideau:
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>> Post images on
http://www.imgur.com and include the link in your posting.
>> *****
>>
>> I've noted the effect Andreas mentions quite often. I usually set the
>> pixel
>> dwell time to the maximum such that the laser will spend the longest time
>> possible scanning out a single line, but even then the flyback can
disturb
>>
>> the reading. The best way is to park the mirrors in the center position,
>> although not all systems allow you to do that. Nikon's old C1 platform
>> allows for it, and ThorLabs' ThorScanLS has a park button as well. I'm
not
>>
>> sure about other vendors, but I'm sure others can chime in with their
>> experiences.
>>
>> Craig
>>
>> On Tue, Jul 18, 2017 at 11:44 AM, Andreas Bruckbauer <
>>
[hidden email]> wrote:
>>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>> Post images on
http://www.imgur.com and include the link in your posting.
>>
>> *****
>>
>> Hi Claire,
>> Confocals usually blank (switch off) the beam on the return and the power
>>
>> meter averages between the on and off phases. Very slow scans are more
>> accurate an I usually use high zoom. Parking the beam is the better
>> option.
>>
>> Best wishes
>>
>> Andreas
>>
>> -----Original Message-----
>> From: "Claire Brown" <
[hidden email]>
>> Sent: 18/07/2017 18:21
>> To: "
[hidden email]" <
[hidden email]>
>>
>> Subject: Re: Assessing phototoxicity in live fluorescence imaging
>>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>> Post images on
http://www.imgur.com and include the link in your posting.
>>
>> *****
>>
>> Thank you for this great article and pointing to many great resources.
>> I wanted to bring up one issue we have had when trying to work on
>> different microscope and compare light density/exposure.
>>
>> For the CLSM microscopes when we use a power meter at the focal plan the
>> power we measure depends a lot on the scan settings.
>> If we park the beam as a point we get one power. If we go to a 100x100
>> pixel array at zoom 1 with a 10x lens the power is different. if we
>> change
>> the scan speed the power is different again. I suspect this is related to
>>
>> how the power meter integrates the light over time and also how sensitive
>>
>> it is spatially across the sensor. We have decide to just quote our power
>>
>> as the power we measure at the power meter with set conditions and we
>> detail those conditions in our materials and methods section of the
>> paper.
>> We try to use a 10x/0.3 planfluar lens with no phase optics if we can.
>>
>> We have stayed away from trying to calculate the power at the sample
>> because a lot of assumptions have to be made. The assumptions may be
>> different for wide-field versus CLSM versus light sheet versus spinning
>> disk and so on.
>>
>> We ran into these issues when just trying to repeat measurements on two
>> different confocals from two different manufacturers. It can really get
>> quite complex.
>>
>> Does anyone have thoughts on this issue? Any cleaver solutions? It is my
>> thought that comparing relative powers on the same instrument is okay but
>>
>> comparing between systems will be very complex.
>>
>> Ideally, it would be good for the manufacturers to have some kind of
>> laser
>> power measurement in the instrument and software that is always
>> monitored.
>> Even if this is just a relative value to the actual power at the sample
>> it
>> would really improve quantitative microscopy and also help in maintenance
>>
>> and trouble shooting equipment. I'm not sure about others but this kind
>> of
>> a feature would really be a strong selling point for me and the core
>> facilities I manage. In many cases these options are already built into
>> the
>> hardware for the service engineers but are not accessible to the end
>> user.
>>
>>
>> Sincerely,
>>
>> Claire
>>
>> --
>> ------------------------------------------------------------
>> Steffen Dietzel, PD Dr. rer. nat
>> Ludwig-Maximilians-Universität München
>> Biomedical Center (BMC)
>> Head of the Core Facility Bioimaging
>>
>> Großhaderner Straße 9
>> D-82152 Planegg-Martinsried
>> Germany
>>
>>
http://www.bioimaging.bmc.med.uni-muenchen.de>> "
>>
>
> --
>
>
> George McNamara, PhD
> Baltimore, MD 21231
>
[hidden email]
>
https://www.linkedin.com/in/georgemcnamara>
https://works.bepress.com/gmcnamara/75 (may need to use Microsoft Edge
or Firefox, rather than Google Chrome)
>
http://www.ncbi.nlm.nih.gov/myncbi/browse/collection/44962650>
http://confocal.jhu.edu (as of May 22, 2017)