Just to be geekey…

Few care to realize that an electron microscope image is a “false image”. Some objects observe what looks like lambertian headlight shading while others exhibit the inverse of that lighting model.

Take a look at the poster boy image of pollen in the wikipedia web site:
http://en.wikipedia.org/wiki/File:Misc_pollen.jpg [en.wikipedia.org]
You'll notice that the smaller pollen in between the larger pollen have bright edges and dark centers. Reverse of default Lambertian shading. No they are not the shape of blood cells but are roughly spherical.

Here's another image where the entire critter is shown in this reverse lambertian like headlight lighting model:
http://farm1.static.flickr.com/232/535424122_ad2369aa79.jpg [farm1.static.flickr.com]

No there are no “lights” in an electron microscope so rim lighting makes no sense to solve this problem.

If you research this further (which all prospective shader writers should do ) you will find that smaller thinner structures will exhibit this reverse-lambertian shading while larger solid objects shade with a lambertian response. Then there is the electron microscope itself. Depending on the model it varies as well. All depends on how the structures were coated with a metal (usually gold as it can be put in an incredibly thin layer) and how the electron microscope renders the collected bounced electrons' energy.

You will want a control on your electron microscope shader that allows you to flip the shading on a per structure basis if you really want to leapfrog the endless supply of physically incorrect electron microscope shaders that seem to proliferate the web-o-sphere…