Mosquito larva

gpmatthews · Post by **gpmatthews** » Sat Jan 27, 2007 10:18 am

From above

View of underside

Both images taken through my Wild M8 stereo zoom microscope. The green tint is due to the colour of the anaesthetic solution I used to immobilise the specim en(a lidocaine throat spray).

Charles Krebs · Post by **Charles Krebs** » Sat Jan 27, 2007 12:07 pm

Hi Graham... nice full body shots. These larva offer a lot to study at different magnification levels. One of the worlds most prolific and deadly (some species, as disease carrier) insects. You've mentioned the lidocaine throat spray a few times. I've carefully checked ingredients here, but I don't think it is available (at least OTC) in the States. On occasion it would be nice to have an effective anesthetic.

When I looked at the shots I wondered how, if desired, one could eliminate the color of the anesthetic liquid and background. The way it varies from green to magenta made me think it wold be tricky. Then I remembered the Olympus Background Subtraction Toolkit.
http://www.olympusmicro.com/primer/digi ... nload.html
I ran the second image through it (and then adjusted levels) and I while not perfect, it really did a pretty good job! It's free and a pretty nice tool to have on hand.

This is the background it generated (to be "subtracted")

This is the image it generated after subtracting the background. The residual green and magenta could probably be cleaned up completely quite easily if desired.

bernhardinho · Post by **bernhardinho** » Sat Jan 27, 2007 12:57 pm

That's it Charles. How about this:

Bernhard

gpmatthews · Post by **gpmatthews** » Sat Jan 27, 2007 2:12 pm

Thanks both... I has forgotten about that tool - it was sitting on my machine all the time, although it refuses to process the full size image. Must remeber to use it more often.

gpmatthews · Post by **gpmatthews** » Sat Jan 27, 2007 3:27 pm

Here is a reworked image: this was performed on the full size master, but as I cannot get the background removal tool to run on a full size file, I generated the background, resized it to match the master and then performed a difference function in PSP. This gave me a negative image which I then inverted to positive. I then filled any off-colour areas with the new background. It seems to work...

beetleman · Post by **beetleman** » Sat Jan 27, 2007 6:05 pm

Wonderful shots Graham. Super nice details in all of them.

Charles Krebs · Post by **Charles Krebs** » Sat Jan 27, 2007 8:47 pm

It seems to work...

I'll say. Looks excellent!

I forgot about the size limitation as well. I actually did the background "resize" bit when I first downloaded the and played around with the Olympus routine. I had come up with a way to "subtract" the resized background in Photoshop... but now I forget how I did it!

discomorphella · Post by **discomorphella** » Sat Jan 27, 2007 11:10 pm

Hi Graham--

Its a great shot of the larva, and using throat spray as an anesthetic is very clever as well (I hope you bought it for the bug and not because you had a sore throat). You must have some tough bugs where you live, they are all hibernating around here. One other background subtraction method you might want to try is the image calculator in imageJ ( http://rsb.info.nih.gov/ij/ ). It routinely handles large tiff files for me at home and work, and you can't beat the price....

--David

svalley · Post by **svalley** » Sun Jan 28, 2007 6:35 pm

Graham, nice shots. Cool idea with the anesthetic.

gpmatthews · Post by **gpmatthews** » Sun Jan 28, 2007 6:44 pm

Lidocaine works quite slowly because it is usually presented as a water soluble hydrochloride, whereas I find benzocaine is faster, being fat soluble (and therefore partitions into certain parts of cells better, particularly crossing cell membranes with a fatty nature), but has the disadvantage that you have to wash the specimen to remove the poorly water soluble crystals that form. The mosquito larva took ages to be immobilised, and even then kept twitching, rendering it impossible to get good image stacks (CombineZ5 worked better than Helicon in the face of various misalignments, however, but none of the results were presentable).