
So in this post, I'm going to actually describe one of the things I do here, since it's so hard to explain without a picture. Above is just a screenshot from one of the desktops that I'm using exclusively to draw the amygdala and hippocampus on the left brain (I'll do the right side later) of some dude who may or may not have VCFS, AKA deletion 22q11.2. >30% of people with VCFS end up with schizophrenia. Hence my lab's interest in the topic.
The boring part: First, I have to use a million different scripts and programs written by the same people that wrote the program you see filling up most of the screen above (Slicer, www.slicer.org). These take the original scan, compress it so our network doesn't overflow, move the different parts around so they are in the same space, segment and measure out the volumes of grey matter, white matter, cerebral spinal fluid and other less relevants stuff. Then we manually edit this to get an accurate measurement of the entire intra-cranial contents, which we will use to make sure that when we compare e.g. hippocampuses between people that we're comparing the volume relative to the brain size. Otherwise, extremely large-headed men would all seem geniuses etc.
Then I realign the image so that the mid-saggital (between the two hemispheres) line is straight coronally (top right and bottom right images at different zoom levels) and axially (top left image), and also that the brain is "straight" between the anterior commisure (AC) and posterior commisure (PC). These are marked for your viewing pleasure with little pink dots on the saggital (bottom left image) view. There's a lot of tricky bullshit that goes on with this concerning voxel (volumetric pixel) dimensions, and the main purpose of what I'm doing now is to figure out the best way to do this for the brain structure I'm measuring.
Then I literally draw the amygdala and hippocampus according to a relatively strict set of rules that have proved reliable and moderately valid over time in our lab. In the top left image you can see an axial slice of the amygdala (in orange) and the hippocampus (in blue). The line between them is arbitrary. They actually overlap, but for reliability purposes we include a little of each in the other. I actually draw it in the coronal view (top and bottom right), where the main instruction is literally to color between the lines (the lines being white matter tracts and CSF).
The fun part:
Eventually, we'll get to measure this and find out if the amygdala and hippocampus in VCFS patients are similarly affected the way schizophrenic patients are, we'll check to see which neuropsych, clinical psych, and psychopharm stuff it correlates with, and we'll also be able to see if this correlates with the 30% of these patients that are schizophrenic. Finally, (I think) we'll get someone else to do microarray analysis of our patients to see which genes correlate with which of these differences. Cool huh?
Also, since I'm the official amygdala-hippocampus drawer, I'm also heavily involved in determining what protocol we will use in this and other studies. The OpenOffice Calc (basically Excel) window to the right is a huge spreadsheet showing all kinds of crude difference measures between different types of drawings. Eventually, this will all get fed to SPSS, a statistics program that will chew it up and spit out the reliability, but until then I can easily modify my technique for the better using mathematical trickery like subtracting the left from the right and comparing percentages.
What you can't see is that, because this computer's moderately bad-ass and it's running Linux, I have four other desktops open, each doing something. This one is checking e-mail, reading today's words of the day, paying my credit card bills, and keeping my Google Calendar. Next door we're listening to Of Montreal
Sunlandic Twins. Next door to that a
perl script that one of the CS people wrote is creating .iso images and creating a detailed log of a backup project for me. etc. etc.
So, that's what I do - stare at a computer screen all day, color pictures, check my e-mail, and run various scripts with various programs. We also have meetings sometimes. And lots of lectures and seminars to help us learn exactly what we're coloring. I'm also database manager, but that's another story.
Any questions, ask my boss, cause that's all I know.