Microscopy: Deconvolve the PSF and enhance resolution

  • Context: Graduate 
  • Thread starter Thread starter eoghan
  • Start date Start date
  • Tags Tags
    Microscopy Resolution
Click For Summary

Discussion Overview

The discussion centers around the limitations of resolution in microscopy, particularly in relation to the point spread function (PSF) and the deconvolution process. Participants explore the implications of the Rayleigh criterion and the challenges associated with enhancing resolution through deconvolution techniques.

Discussion Character

  • Technical explanation
  • Debate/contested
  • Mathematical reasoning

Main Points Raised

  • Some participants assert that deconvolving the observed image can eliminate diffraction effects, potentially allowing for the resolution of closely spaced points, but question whether this understanding is complete.
  • Another participant references a passage from a book discussing the limitations of the Rayleigh resolution criterion, particularly in high-resolution electron microscopy, and seeks clarification on why suitable experiments cannot be conducted under this criterion.
  • Concerns are raised about the ill-posed nature of the deconvolution problem, indicating that solutions may not be unique or may be corrupted by noise, which complicates the enhancement of resolution.
  • It is noted that while deconvolution is a standard technique for improving resolution, it has inherent limitations that prevent the complete elimination of diffraction or aberrations.
  • Participants agree that approximate deconvolution methods are commonly used, but emphasize that mathematical principles explain why diffraction cannot be entirely removed.

Areas of Agreement / Disagreement

Participants express a mix of agreement and disagreement regarding the effectiveness and limitations of deconvolution techniques in microscopy. While there is acknowledgment of the utility of these methods, the discussion reveals ongoing uncertainty about their capabilities and the implications of the Rayleigh criterion.

Contextual Notes

The discussion highlights the ill-posed nature of the deconvolution problem and its susceptibility to noise, which are critical factors in the context of enhancing resolution in microscopy. Limitations related to the Rayleigh criterion in high-resolution experiments are also noted but remain unresolved.

eoghan
Messages
201
Reaction score
7
Hi all,

In a diffracted limited microscope the resolution has a limit given by the Airy disk and this gives rise to the Rayleigh criteria. But if I deconvolve the observed image removing the psf then I eliminate the diffraction effects and I can resolve any distinct points, whatever close they are. Am I missing something?
 
Science news on Phys.org
Also, another question related to the previous one. In the book Three-dimensional electron microscopy of macromolecular assemblies by Frank, ed 2006 ch 2.5.1 - "Concept of resolution" it is said

"Optical definitions of resolution are based on the ability of an instrument to resolve two points separated by a given distance d. [...] [Another] criticism is that the resolution criterion formulated above does not lend itself to a suitable experiment when applied to high-resolution EM. On the nanometer scale, any test object as well as the support it must be placed on, reveals its atomic makeup."

What does this passage mean? Why cannot have suitable experiments with the Rayleigh resolution criterion?
 
eoghan said:
In a diffracted limited microscope the resolution has a limit given by the Airy disk and this gives rise to the Rayleigh criteria. But if I deconvolve the observed image removing the psf then I eliminate the diffraction effects and I can resolve any distinct points, whatever close they are. Am I missing something?
The problem is that the deconvolution problem is ill-posed, meaning that a solution may not be unique, or may not even exist, and/or is quickly corrupted by the presence of noise, no matter how tiny.
 
olivermsun said:
The problem is that the deconvolution problem is ill-posed, meaning that a solution may not be unique, or may not even exist, and/or is quickly corrupted by the presence of noise, no matter how tiny.
Nevertheless deconvolution is nowadays a standard technique to enhance resolution, though, for the reasons mentioned by you, it has it's own limitations.
 
Sure, approximate deconvolution methods are used all the time to improve images. But the mathematics tell us why we can't just eliminate diffraction or aberrations once and for all and be done with it. ;)
 

Similar threads

  • · Replies 13 ·
Replies
13
Views
3K
  • · Replies 10 ·
Replies
10
Views
7K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 11 ·
Replies
11
Views
4K
Replies
2
Views
5K