What lies beneath... or within... or something

Magnetic Resonance Imaging, or MRI, is a imaging technique used in medicine to visualize internal structures of the body in detail. MRI makes use of nuclear magnetic resonance (NMR), a technique beloved of physical chemists and biochemists to help them understand molecular structures, to image the nuclei of atoms within the body and build up a 3D picture of internal structure. This is particularly useful when looking for gross differences in organs or other body componentry associated with disease. ('Gross' meaning outward anatomical description, not as in 'fuck that dude's bowel looks gross', although these definitions can overlap in clinical presentation depending on how professional your radiologist is.)

Of course, you don't have to use MRI for medicine. Pretty much anything can go into a MRI machine. Even the contents of your vegetable crisper, as per this bloke at Boston U with a lot of spare time on his hands and on his instrument booking sheet.

Orange you glad this is from a dedicated research instrument and someone's not waiting an extra week to have their brain tumour imaged

This, not to put too fine a point on it, is fucking cool. I think we in the life sciences (the Royal 'we' meaning everyone else who still has a job in the life sciences), who are very familiar with techniques like this, or at least seeing the results of them, forget that being able to see inside stuff which is inherently not-see-inside-able by design, is often found to be immensely fucking ace by the majority of non-SCIENTS types. We can get a bit blase about 3D rendered reconstructions of MRI or confocal microscope images spinning on some droning dullard's conference Powerpoint slide. Stuff like the above - while utterly pointless and almost definitely a waste of profoundly expensive instrument time - has a massive capacity to capture the imagination of Normal Humans and should never be underestimated.

I mentioned above that MRI data isn't the only way you can 'see through' solid biological objects. Confocal microscopy, which is a big-shiny-expensive way of imaging samples with fluorescent labels introduced into the sample, also uses the same idea - that of taking multiple 'optical sections' through a sample, sometimes called 'Z-stacking' (i.e. taking lots of 2D images at varying points through the imaginary Z-axis of the sample, then computationally stacking them all on top of each other.) You can even do this with high-powered stereomicroscopes - the sort you'd use to look at larger solid objects like insects at high magnification - with an electronic motor that can 'step' the focus drive through the image, to capture images with a CCD camera at each different stage, then piece a completely-in-focus high-magnification version of the object together. The complication here is that the higher magnification you use, the shallower your depth of field becomes - or in other words, more of the overall beastie will remain out of focus. Same is the case with 'non-confocal' fluorescent microscopes. Confocals are clever because they can effectively 'blank out' all the out-of-focus light information which comes bouncing back to the viewer when the fluorophores in the sample are illuminated.

What all these techniques depend absolutely on, therefore, is deconvolution software which can piece together a coherent, in-focus 3D image from partially-focused 2D Z-stack images, those 'optical sections' I described above. (Compare this idea to an 'actual' tissue section - for instance what goes on in a pathology lab where a piece of tumour sample gets 'fixed', embedded and sliced very thinly for staining on a histology slide and examination by a pathologist. You could theoretically recreate the 3D structure of the tissue sample by 'serial' sectioning and staining a series of sequential slices... but it'd suck. And this is a cleverer and easier way to do it.) Obviously, the ability of the software to take on board all the masses of information it gets from a huge Z-stack of high-res images, identify the out-of-focus information, and discard it so sense can be made and a coherent overall picture can be formed, is the most critical part of the whole process.

And it struck me while I was contemplating all of this, that this is pretty much what I've had to train my brain to do as a means of counteracting anxiety.

I've mentioned before my run-ins with the Despair Squid. What anxiety is (for me at least) is a massive unwanted oversupply of unhelpful thoughts effectively crashing the system through a monumental DDoS. Everyone's different, everyone's triggers vary and everyone's techniques for dealing with are going to be very individual, but the inability to slow down the thinking process, challenge 'catastrophisation' (a word I really like) and reintroduce some rationality to the discussion, usually correlates with Despair Squid Wrestling. So this idea of taking one's thoughts apart, top to bottom, piece by piece - optically sectioning them if you like - then analysing each of them, challenging the data, throwing out the out-of-focus noise and retaining the useful stuff in order to piece together a more coherent view of the situation - that's actually one of the more apposite analogies I've found for my particular process of seeing off anxiety.

I have no idea whether this will be useful information but I thought I'd share it with you.

Flangebadger.

The Doctor is OUT.