John,
Let me start by saying that I think this is a great thing to include in
imageJ and for the light/fluorescence community to consider as an
analysis tool.
I come from the electron microscopy/tomography community and my other 2
cents is a reminder that above all else, the Fourier Ring Correlation
(FRC) measures the similarity between two images as a resolution
dependent function. Whether that relates to what anyone means by
"resolution" is a completely different matter.
In addition, there are discussions among a number of very smart people
in the EM community about what numerical value of the FRC indicates a
"significant similarity" (and thus to what resolution a pair of images
are similar). These discussions at times can take on the airs of
religious certainty and turn into rather nasty arguments.
I don't think any of this detracts from the utility of the FRC, but it
does mean that people who use it should understand what they are
measuring and what assumptions they are making, and to take the results
with a least a small grain of salt!
On 05/27/2013 05:07 PM, John Oreopoulos wrote:
> I would like to bring to the attention of the ImageJ community a very important (in my opinion) research article which is now available on the advanced online publication section of Nature Methods:
>
> Nieuwenhuizen, R.P.J., et al., Measuring image resolution in optical nanoscopy. Nat Meth, 2013. advance online publication.
>
>
http://www.nature.com/nmeth/journal/vaop/ncurrent/abs/nmeth.2448.html>
> This article presents the concept "Fourier ring correlation" - originally developed for electron microscopy/tomography of single biological protein complexes, now applied to optical super-resolution localization microscopy - a method which, among other things, can directly calculate the image resolution from the raw image data without relying on any models or external/indirect calibration. The method can also be used to determine when is a good time to end a super-resolution localization microscopy acquisition session (ie: identifying the frame number at which point the resolution is not improving anymore). Thankfully the authors of this paper provide an ImageJ and Matlab plugin which can be used to apply the calculation to localization microscopy data points (a list of xy localization coordinates). The authors also point out that this method of figuring out the image resolution can be applied to other diffraction-limited forms of fluorescence imaging, including TIRF and c!
onfocal mi
croscopy. All that is required in this case is two or more images of the same sample that only differ in noise content. While the Matlab routine can be used with regular fluorescence imaging data, unfortunately the authors have not had the time to extend the ImageJ plugin in the same manner.
>
> I see value in the idea that Fourier ring correlation could be used as a model-independent measure of microscope image resolution across all forms of fluorescence imaging (super-resolution or diffraction-limited), and if it were available to everyone through ImageJ, perhaps it might become a more accepted standard for reporting image resolution. I am wondering if there is anyone out there in the ImageJ community able and willing to take on the challenge of extending the authors ImageJ plugin for this purpose. This task is beyond my coding capabilities.
>
> Sincerely,
>
> John Oreopoulos
> Staff Scientist
> Spectral Applied Research
> Richmond Hill, Ontario
> Canada
> www.spectral.ca
> --
> ImageJ mailing list:
http://imagej.nih.gov/ij/list.html>
--
David Gene Morgan
cryoEM Facility
320C Simon Hall
Indiana University Bloomington
812 856 1457 (office)
812 856 3221 (EM lab)
http://bio.indiana.edu/~cryo--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html
Locked
May 27, 2013; 9:07pm
