Nanoscopy awarded Nobel Prize for 2014

[hkv]

Great news and an interesting choice of chemistry award winners

"Surpassing the limitations of the light microscope"

http://www.nobelprize.org/nobel_prizes/ ... press.html

[Beatsy]

Neat stuff!

I've read three press releases that claim they can "...exceed the abbe limit for resolution". Is that actually true? Can they resolve two molecules that are less then the abbe limit apart? Or are they only detecting fluorescence from single molecules at a time?

[Pau]

I've read three press releases that claim they can "...exceed the abbe limit for resolution". Is that actually true? Can they resolve two molecules that are less then the abbe limit apart? Or are they only detecting fluorescence from single molecules at a time?

Interesting question, as far as I understand what they call "nanoscope" is usually called superresolution, and yes, that systems are able to resolve details well beyond the venerable Abbe's limit. The tric is the use they do of the fluorescence signal, easier to understand with the single molecule method: two molecules fluorescing toghether at less distance than the Abbe's limit will not be distinguishable because the diffraction disc but if they are fluorescing in different moments/focal planes the software can reduce its dimension to small points and therefore resolve them. (maybe not a good explanation, just my superficial understanding)

[TheLostVertex]

Thanks for this. The pdf in your link ( http://www.nobelprize.org/nobel_prizes/ ... ze2014.pdf ) on the background is a very good and interesting read.

Edit:

The tric is the use they do of the fluorescence signal, easier to understand with the single molecule method: two molecules fluorescing toghether at less distance than the Abbe's limit will not be distinguishable because the diffraction disc but if they are fluorescing in different moments/focal planes the software can reduce its dimension to small points and therefore resolve them. (maybe not a good explanation, just my superficial understanding)

This is my understanding as well. When the distance between two flourophors is larger than Abbe's limit they can be resolved at a better than diffraction resolution. The way they it gets resolved depends on the method used. One method uses lasers to make the spot size smaller and scans the object. The other method excites samples that are far from each other and uses their PSF to find their center; it does this many times and then merges all the collected samples together to form an image(in a way that appears ALMOST analogous to image stacking).

[Beatsy]

Thanks both. That paper is an interesting read and has helped a lot in understanding what's going on.

Abbes limit is being exceeded, but only in terms of the fluorescing spot-size. Normally, you can't focus a laser to a spot smaller than 1/2 a wavelength (due to Abbes limit), but they extinguish 'extraneous' fluorescence to subsequently shrink the 'excited' spot well below Abbes limit. Then (from the paper linked above)...

The knowledge that there is a single emitter allows the position of the point source to be estimated much more precisely (c) than the width of the PSF (b).

Which makes sense to me now. They can only pin down a single point source to an accuracy less than Abbes, but they could not resolve two point sources to the same degree of resolution.

In summary, Abbes is being exceeded in terms of the spot size fluorescing, but is not being exceeded in terms of (simultaneously) resolving two spots that are <0.5 of a wavelength apart.

I think...

[TheLostVertex]

In summary, Abbes is being exceeded in terms of the spot size fluorescing, but is not being exceeded in terms of (simultaneously) resolving two spots that are <0.5 of a wavelength apart.

I think...

That is what I get out of it. They are just resolving single spots and then using those spots to form an image of the object.

[jswatts]

Hi, guys,

I work at a microscopy facility at a large research university and have been following super-res out of professional interest. Here are my two cents, if anyone is interested.

There are basically three methods used for achieving images that break the Abbe limit. The first (the one you are discussing) was developed by Eric Betzig, Jennifer Waters and Mike Davidson and it uses the flickering behaviour of some fluorescent molecules to gather information about their location. A small subset of the population of molecules is turned on, and numerous images are taken of them, each image being a normal, Abbe limited image. A new set of molecules is turned on and the process repeated. About 10,000 images are acquired in all. For each molecule, you then have a whole set of measurements of the diffraction disk from which the location of the centre of the disk can be found; this gives the location of the molecule. The size of the spot on the image is given by the standard deviation of the location of the Airy disks associated with that molecule. These images are best understood as statistical representations of image data. This is the system used by Nikon.

The next type of super-res was developed by Stefan Hell, and uses an intense ring of emission frequency light to cause fluorescent molecules to emit their light away from the objective (kind of like a laser). The fluorescent molecules in the centre of the ring can emit their light in any direction. The size of the ring, and the centre of the ring, can be adjusted so that the detectable fluorescence comes from a spot smaller than the classical diffraction limit. This is the system used by Leica.

The third method (which I don't understand so well) projects an interference pattern (a stripe pattern dependent on the wavelength of the excitation light) onto the sample. This pattern is rotated, and the high frequency information is extracted using analysis software (using Fourier analysis).

I hope this helps a bit. As previous posters have noted, these methods only work for fluorescence imaging. There are some other methods, like near-field scanning optical microscopy, but they aren't as useful when dealing with living subjects.

Cheers,
John

[rjlittlefield]

The third method (which I don't understand so well) projects an interference pattern (a stripe pattern dependent on the wavelength of the excitation light) onto the sample. This pattern is rotated, and the high frequency information is extracted using analysis software (using Fourier analysis).

This sounds like the method discussed HERE. Unlike the other two, it can be made to work for non-fluorescent subjects also. But also unlike the other two, the maximum improvement in resolution is limited to 2X or so.

--Rik

[jswatts]

Hi, Rik,

Yep, the third method is SIM - structured illumination microscopy.

Cheers,
John

[l2oBiN]

http://www.leica-microsystems.com/scien ... esolution/