Microscopy: Deconvolution Microscopy (David Agard)

www.ibiology.org/talks/deconv...
Deconvolution is a technique to calculate a model for the object that gave rise to the microscope image using knowledge about the way the microscope forms images. This lecture describes the basic principles of various deconvolution techniques and introduces principles important to deconvolution such as the Fourier transform, points spread function and optical transfer function.

[@zhenhongdu]

I took about 3 hours to carefully listen to this great talk!Maybe this is what the master like!

[@liangwenxuan]

The correct grammar is: it took me 3 hours to ...

[@andreamckerry2961]

This is such a great video. Not only does it tell you what it is but gives a good scope of why it is useful and applications to use it for.

[@pianofantast91]

You are the best teacher ever.

[@NguyetMinhTDAO]

Thank you so much for this video!

[@leighstanger8988]

great video thank you. I wish i had seen it before i stared writing my thesis like my ideas were not already well known.

[@JulioSSierraCamarena]

Nice explanation, clear and detailed.

[@davidskeet6220]

Thank you for describing this very powerful technique. I am an amateur microscopist and would like to try the simplest 3-stack, "nearest neighbour" deconvolution technique using Matlab on acquired images. But obviously, before I start, I need to acquire the PSF of my optics. You didn't mention how PSF's are acquired. Is there a special test slide that can be purchased or constructed? Presumably it would be a grid of small apertures in an opaque thin material?

[@SahinKupusoglu]

Great video, thank you...

[@moniii_666]

This is sooooooooo helpful. Thank you

[@user-eu5lw5rs3p]

Very useful, thanks a lot

[@lavandeneth9965]

watching a master level Video from undergraduate level ;)

[@yn8093]

Thank you for this video. 🙆

[@lingyunwang3951]

It's so good, and it can help me learn more about PSF. Actually, i do some work about it.

[@stevengill1736]

Deconvolution - great name for a punk band! ;*]
This reminds me of how telescopes correct astronomical images - neat!

[@blanca4287]

Could you explain the difference between deconvolution and extended depth of focus? Thank you.

[@memyselfpearl]

Using deconvolution, you remove the blur computationally with the help of PSF and thereby improve the spatial resolution. This can also be applied for a 3D stack. In the extended depth of focus, your images are still blurred and the signal to noise ration could be still sub-optimal. Hope this helps.

[@cryora]

@@memyselfpearl What is extended depth of focus?

[@shirazkaderuppan3279]

This is a good question. To answer your question, in deconvolution microscopy, we seek to obtain spatial information which has been obscured due to the haze/blurring distortions introduced by the optics of the microscope (termed the PSF). So by means of deconvolution, you gain additional insights from the raw image, as you can now resolve additional detail within it (which cannot be obtained directly from the raw image). In extended depth of focus/field (or EDF), we simply combine different in-focus regions of an image (acquired at different image planes/sections) to generate an overall in-focus (HDR) image. No additional information may be gleaned by EDF, as the distortion introduced by the PSF of the optical system is still present (an in-focus region is still subjected to optical distortions introduced by the PSF of the system).

[@lpjunction]

The same technique should be applied to images of Jupiter and Mars.

[@BrandonSLockey]

exam gods plzzzz give good grades so i can do systems bio with my friend (T_T)