Posted by
Barbara Foster on
URL: http://confocal-microscopy-list.275.s1.nabble.com/DIC-or-Phase-contrast-for-image-morphometry-mesasurements-tp7068286p7071255.html
*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy*****
Hi, Monique
Several basic fundamentals to consider:
a. RE: Phase
Phase suffers from haloing, an artifact which can obscure edges.
Having said that, the technique can be optimized in two ways. Both
depend on the fact that Phase is engineered to work best when there
is a lambda/4 shift in light passing through the mountant
("background) versus the light passing through the sample. The
further you deviate from this engineering requirement, the worse the
halo will become. Without going into all the physics, as to the
source of this artifact, here are the practicalities: you can
optimize the RI match between the sample and the mountant AND you can
optimize the wavelength (lambda). Since you are working with a living
organism, you will be restricted on what mountant will work. Water
(RI 1.33) is probably preferable, but you might want to also try
something like a glucose solution (Caution: yeast love sugar and will
produce CO2 which, in small quantities, should just dissolve, but in
larger quantities will cause bubbles in your sample). On the
wavelength side, your phase kit should have come with a green
filter. Check with your microscope vendor as to which wavelength
(typically 546nm or 589nm) your phase kit was engineered to use. If
the filter is just a glass filter, you can improve the situation by
using a narrow band pass interference filter. Here again, there will
be a trade-off... you may need more light going in and that might not
be good for your yeast.
b. RE: Darkfield
Two drawbacks here:
Darkfield has infinitely deep depth of field, so if your preps are
not really clean, you may get interference from objects above and
below the real plane of focus.
Depending on the diffraction at the edge, you may also get thicker
boundaries, making exact location of the edge difficult to
determine. Again, you might be able to optimize by playing with the
mounting medium.
c. RE: DIC
I agree with the earlier posting that the bright edge/dark edge can
be a problem, but as I understand it, some of the newer software
programs have algorithms to address this issue. It's been a while
since I worked in this area, but I seem to remember edge filters
(Sobel? Roberts?) which might also help. Alternatively, if your DIC
is tunable to the point that you can change the background from
"first order gray" to "first order red", you can get away from the
bright/dark problem and just get "color contrast" (one edge blue, one
edge yellow). Because DIC is typically used to detect gradients and
the most sensitive setting is when the background is dove gray, we
typically don't advise folks to work in this range, but it is
certainly doable. Of all the techniques, this one will provide the
best, crispest edge definition and the most shallow depth of field
(tunable using your aperture iris). For all the details as to why
this works, see my article in American Lab from April 1988. (I can
email you a PDF, although it is not a very good copy).
Good hunting... and let us know how you make out!
Best regards,
Barbara Foster, President and Sr. Consultant
Microscopy/Microscopy Education
P: (972)924-5310
W: www.MicroscopyEducation.com
We are now scheduling courses through June 2012
At 09:33 AM 12/7/2011, you wrote:
>*****
>To join, leave or search the confocal microscopy listserv, go to:
>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>*****
>
>Have you test any color camera to do that ? May be it is possible to do that
>whit a bw camera with a green filter. Most of the time the filter increase
>contrast !
>At yet i measure colonies on agarose substract and it's realy easy to do that.
>I'm using MetaMorph software and make threshold using size, circularity,
>shape and any filter you want...