Re: History lesson

[Zdenek Svindrych]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Andrew,

Indeed, your post is essential to understanding the Wikipedia article... The
wiki page is really confusing, perhaps due to the unfortunate fact, that the
valuable Christoph's work somehow stayed off the main stream of the
superresolution microscopy and the terminology has evolved somewhat.

Just a note (I can't resist): the resolution you routinely achieve (1 to 2
nm) is astonishing!... We routinely achieve XY drifts of 5 nm per minute...

Cheers,

deden


On Wed, 5 Dec 2012 12:01:43 -0800, David Baddeley
<[hidden email]> wrote:

>having done my PhD in Christoph's lab I can confirm that he
>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Hi Andrew,

having done my PhD in Christoph's lab I can confirm that he has a
considerable pedigree in the fields of high and super- resolution microscopy
and is probably not cited as much as he deserves. I can also confirm that
all the claims on the wikipedia page will have some basis in measurement.
That said, the wiki page is somewhat confusing and arguably conflates
results from multiple different microscopy techniques (or fails to draw some
of the distinctions/ offer some of the explanation that would be required to
fully understand what is being claimed). Part of the reason for this is
probably that the page was prepared by a member of the universities
intellectual property team (judging by the edit history) who might not have
fully appreciated some of the subtleties. Christoph himself can arguably
also sometimes be guilty of taking a slightly optimistic view whilst trying
to paint a broad picture. 

Addressing the two topics you have picked up on:
- 30-40 nm structured illumination resolution. This is size resolution
and/or position resolution, not spatial resolution. The SMI microscope
generates an axial standing wave between two microscope objectives,
resulting in a structured illumination pattern with ~ 160 nm period. When an
object is moved through this pattern, the resulting signal will have a
modulation, the magnitude of which depends on it's size (infinitely small
objects will dim completely in the troughs, whereas objects > ~ 200 nm show
very little modulation. Based on the modulation observed, an estimate of the
real size can be obtained. On objects with a well characterised shape (such
as beads) it is possible to get an accuracy better than ~10 nm, but the
method becomes less accurate when assumptions need to be made about the
underlying object. 30-40 nm is probably a reasonable estimate of the
precision with which sizes of biological structures can be determined. When
determining
> positions, the illumination pattern can be used as a 'ruler' to improve
axial localisation (not too dissimilar to an iPALM like approach). For
bright objects (not single fluorphores), I regularly got 1-2 nm localisation
precisions.

The references in the SMI section are not particularly helpful, in that they
refer to two separate techniques - some (9, 12) are conventional lateral
structured illumination, whereas 10 & 11 refer to the nano-sizing technique.

- 10 nm resolution in 2 minutes. It is not completely clear whether this is
referring to a PALM/STORM type of approach, or an SMI type approach. SMI
only needs about 20 frames to scan the object through the focus, so this is
easily achievable. I suspect that it however refers to PALM/STORM type
imaging and that the 10 nm is a localisation precision rather than a nyquist
sampled end resolution, in which case it could possibly be considered
oversold. It might well be a conflation of both.


One source of confusion is probably the re-purposing of acronyms (in this
he's following the lead of the the Hell group who have, e.g., changed GSD
from meaning a STED like point-scanning microscopy to a PALM/STORM variant).
Maybe the following (somewhat subjective) disambiguation might help:

SMI - axial structured illumination used for size and/or position
measurements with sub-diffraction (few nm on synthetic objects) accuracy.
Also sometimes used to mean conventional SIM/PIM with a lateral illumination
pattern and a 2-fold resolution improvement. It should be noted that
Christoph was involved in some of the first lateral SIM experiments
(Heintzmann and Cremer, 1999).

Vertico-SMI - an SMI microscope in which the two objectives are orientated
vertically, and with an incubation chamber used for live cell imaging (the
initial SMI prototypes were built on a breadboard with horizontal
objectives). Being a microscope equipped with both a sensitive camera and
laser illumination it was/is also used for PALM/STORM type measurements.

SPDM - a method of sub-resolution structure analysis based on measuring the
relative positions of objects in different spectral channels (e.g. looking
at the relative position of different parts of a gene locus - Esa et Al,
JoM, 2000). Severely limited by the number of available spectral channels,
but conceptually similar to PALM/STORM, and arguably a fore-runner. Also
used / repurposed to mean PALM/STORM

SPDM-phymod - PALM/STORM, following the rational that switching is
conceptually equivalent a spectral signature. Sometimes used to mean
PALM/STORM with conventional fluorophores (we were the first to publish
localisation microscopy images based on conventional fluorophore switching
in Reymann et al, Chromosome research, 2008 - they were included as an
afterthought so you need to dig through the article to find them, and were
by all admissions pretty unconvincing with the switching chemistries etc ...
not yet worked out).

LIMON - catch all acronym to refer to the Cremer group's super-resolution
efforts. Similar in concept to RESOLFT.

cheers,
David