optimisationsTriangle Ever find yourself wishing for the last microscope you will ever need to buy, the instrument that can view anything at any scale and any speed? It’s very tempting to imagine an optical microscope with the diffraction-unlimited resolution of STED, the volumetric imaging speed of light sheet illumination, the deep-tissue penetration of multiphoton microscopy, and the ability to do it all in phase and scattering without invoking a need for exogenous fluorophores or dyes. Perhaps the gamma-ray microscope employed in a Heisenberg thought experiment or the tricorder from Star Trek would come close, but unfortunately we are still waiting on the underlying technologies to mature for the latter. In microscopy as in life, optimisation in one capability comes at a trade-off cost in another. Put more plainly, TANSTAAFL.

Earlier in the summer I attended a biophotonics summer school at the University of Illinois at Urbana-Champaign’s Beckman Institute (link). At the end of a combination of lab tours, seminar-style lectures and poster sessions, we were treated to an hour-long presentation by the president of Carl Zeiss, James Sharp. Perhaps you have heard of Zeiss, a company eponymous with its founder, who teamed up with Ernst Abbe in the late 1800s to invent and commercialise the field of microscopy. After painting a stark contrast of the present job market with that of days past with a story of being stuck in his interviewer’s office for a day by a locked filing cabinet and errant bell-bottom pants, Sharp went on to give what essentially amounted to a 45 minute advertisement for Zeiss (spoiler, they are not best friends with Leica) as a company to work for or buy things from. It was an insightful set of slides that emphasised how far I have to go in my own career before I could fathom spending half a million dollars on a microscope. One insight that I came away with that will stick with me for the foreseeable future is the imaging optimisation triangle.

Sharp described the triangle as a trade-off for traits of resolution, speed, and depth, but the concept is fairly common and the third trait is often defined by the signal to noise ratio, or sensitivity. The moral of the story is that all three corners of the triangle can’t be optimised simultaneously. All else being equal, STED can’t be as fast as wide field or light sheet imaging, and nothing can penetrate tissue like 4-photon imaging. Step changes in the underlying technology can raise the watershed for performance across microscope modalities, e.g. new sensor paradigms can improve signal-to-noise regardless of the technique used. However, even with marked leaps in innovation, you can’t have it all at once.

The microscopy triangle is typically invoked as a qualitative example of trade-offs. However, the three traits certainly have measurable performance features and three corners equates easily enough to three axes. Why not populate a quantitative volume to show the pros and cons of various imaging modalities? Here are a few flagship microscope techniques populated on the quantitative microscopy TANSTAAFL pyramid.


These are all vastly different techniques, so the minutiae of their strengths is somewhat lost. Given a known volume occupying desirable specifications for testing a hypothesis, the graph could be populated with the techniques at your disposal and used to inform a decision on which to utilise. More realistically, axes can be added as need (e.g. for photobleaching, axial resolution) and a single or set of similar techniques could be considered with different settings, e.g. laser power, sensor used, etc., rather than comparing these vastly different modalities.

Values are approximate and from the following sources:
Multiphoton depth penetration extimated from a talk by Chris Xu of Cornell University http://www.jneurosci.org/content/30/28/9341.short
Wide-field: Personal estimates

The triangle is known by various names including, imaging triangle or (somewhat ominously) as the eternal triangle, or even the “triangle of frustration.”

Update 2014/07/09: Typo: “Start Trek” corrected to “Star Trek”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.