A virtual camera lucida

Sometimes, the scientific literature sucks at getting information to you.

I was looking at the table of contents of a new issue of The Journal of Crustacean Biology and saw an article about how to photograph soft-bodied crustaceans. Hm, I wonder why photographing soft-bodied crustaceans is difficult, I thought.

And the abstract mentioned software to deal with short focal planes by merging several pictures. The software is Helicon Focus.

Yes, I should be happy that I have found something useful, but... dagnabbit, why didn’t I know about this years ago?

For instance, here’s part of Figure 3 from a paper a few years back (Espinoza et al. 2006) on the left, compared to a new picture processed with Helicon Focus (click to enlarge).

These are neurons stained through a technique called backfilling. The best backfills are really superb, and you can see a lot of detail. But they’re often in thick tissue, and neurons go all over the place through it, making it hard to see all the relevant detail. The way people usually got around this was to get a camera lucida attached to a drawing tube through a microscope, then trace the axons as you focused up and down.

I was actually very pleased with the pictures in Figure 3 as published, which had dark cells against a very clear background. But the one shown here on the left shows the problem of getting all the relevant detail in the shot. The cell bodies on the right are okay, but the ones down a little deeper in the ganglion on the left and their axons are already blurring out.

The processed version on the right is better.

Having played with this a bit, sometimes there are things you can see in the individual frames that do get lost in the processed version. The biggest problem is when there is some detail underneath something else. You can often see it under the scope and in the individual frames, but not so well in the composite image.

This was so startling and so useful to me, I briefly entertained the thought of trying to turn this into a neuroscience methods paper. Then I looked and found short references that it had been used in the invertebrate neurobiology literature a couple of years ago (Scanell et al. 2008) in a journal I read. D’oh! Another shrewd Faulkes scheme bites the dust.

But at least I can spread the word through a blog post. A slightly more recent paper by Berejnov and company (2009) has nothing whatsoever to do with biology, but gives a better example of what you might get from this kind of software.

As I was writing this, I read an interview with Neil Gaiman, where he said this, mostly in relation to book publishing in particular genres:

Information used to be gold: hard to find, expensive, the equivalent of going off into the desert and coming back with a perfect lump of gold. Now, it’s the equivalent of going off into the jungle, in which there is information everywhere and what you are trying to find is the piece that is useful, while ignoring the noise.

I do wish the process of finding useful things was a little less jungle-like.

References

Berejnov, V., Sinton, D., & Djilali, N. (2010). Structure of porous electrodes in polymer electrolyte membrane fuel cells: An optical reconstruction technique Journal of Power Sources 195 (7), 1936-1939. DOI: 10.1016/j.jpowsour.2009.10.050

Espinoza SY, Breen L, Varghese N, Faulkes Z. 2006. Loss of escape-related giant neurons in a spiny lobster, Panulirus argus. The Biological Bulletin 211: 223-231. http://www.biolbull.org/cgi/content/abstract/211/3/223

Hegna, T. (2010). Photography of Soft-Bodied Crustaceans via Drying, Whitening, and Splicing Journal of Crustacean Biology 30(3): 351-356. DOI: 10.1651/09-3253.1

Scannell, E., Dell'Ova, C., Quinlan, E., Murphy, A., & Kleckner, N. (2008). Pharmacology of ionotropic and metabotropic glutamate receptors on neurons involved in feeding behavior in the pond snail, Helisoma trivolvis Journal of Experimental Biology 211(5): 824-833. DOI: 10.1242/jeb.011866