Tuesday 28 June 2016

Anti-scientific bias of 'scientific' animations

How does personalising cells destroy the magic?

A review of anatomy and physiology videos


Naturally, blood is typically represented as red, as here. The function of Red Cells [Anatomy World]. Red cells are enlarged and spread out to make them visible. We are used to such distortions to make cells visible; perhaps you think I'm being pedantic but a dilution of blood cells like this would be incompatible with life. Anyway, depth is implied by showing artificial incident light and variation in scale and usually the cells are rushing past us to suggest cells propelled by the pumping of the heart and suspended in plasma.


In this video, Red cells are given human features in order to describe: A Day in the Life of red blood cell [Digital disasters] . Red is the obvious colour although in the absence of light, ie inside the body, blood is black but so is everything else. 
Each red cell is given a face, a personality and a voice.
To me, giving faces to organs and cells creates a childish, fairy tale atmosphere which is patronising, deceptive and inaccurate.
If we personalise our organs, we overlook the way they can sometimes fight each other, harbour cancer or weather faster than we would like.  

Darwinian evolution implies that there is no organising intelligence, just an incredible interplay of ultimately autonomous physiological processes which sustain life by constantly maintaining temperature, oxygen, acidity and glucose within a narrow range to sustain the miracle of life. 

Here's a contrasting film: Haematology - Red Blod Cells [Armando Hasudungan]
Perhaps this errs on the opposite end of the scale: the microscopic laboratory vision with conventional ignoring of scale. The danger here is that anatomy is only part of the picture. At heart it is a question of what we want people, medical students, doctors to know of all the facts that have been discovered about blood. This video poses as a scientific account but is only a partial view.

One variant of the red video is in fuchsia pink with contrasting light blue, here used to show the adhesion of platelets to a clot.


Blood, Part 1 - True Blood: Crash Course A&P #29 [CrashCourse]
Blood vessels are usually shown as diagrammatic cylinders, as here. Again, red cells are massively enlarged and widely spaced out. Even this largely 'scientific' film veers into characterisation of whicte blood cells as people:




 This feels sentimental and old-fashioned. Can the audience bear to live in an existential world where our function depends on ingrained, automated reactions to random variation? There is a creepy angle to personalising cells: what if, like people, they become argumentative, temperamental, self-destructive? But this unsettling option is rarely evoked. Most videos would like us to bathe in the comforting image of our tissues and cells constantly working lovingly to help us live our lives with a smiley face.


Sunday 26 June 2016

How medical illustration distorts the facts

How medical illustration distorts the facts

As soon as art is used to explain science a rift opens up between reality and the image.  

This happens in human biology for a variety of reasons.
  1. Biological events happen in complete darkness: hormone secretion, clot formation and the pumping of the heart happen deep inside the body where there is no light and no convenient gaps between the organs. The heart is wrapped around by the lungs, the pancreas is beside the liver and the duodenum. In order to see the actions of a single organ you need to isolate it from its surroundings but the movement of the lungs influences the emptying of the heart. Even an endoscope inside a blood vessel cannot see the vessel wall unless the blood is flushed out of the system. That's why radiologists use contrast medium to create a 3D image of the circulation.
  2. Most biological events are too small to see with the naked eye or the optical microscope, so we need to construct an image to describe what is happening on a scale of nanometers [one millionth of a meter] or molecular events at the level of picometers, maybe sometimes at a sub-atomic layer in the case of changes in quantum physics which occur inside the body. 

  3. When events like clotting are scaled up so we can observe them there are deliberate mistakes in scale; to show platelets and the vessel wall in the same image, one molecule of haemoglobin is made to look huge in comparison to one red cell.

  4. Biological events, like the pumping of the heart are dynamic, changing in a rhythmic cycle over seconds, hours and a whole day. So to capture what's happening you can opt for a snapshot like a sagittal CT scan or use ultrasound to record an echocardiogram over several seconds to observe dynamic emptying and filling. Either way, it's a compromise between accurate anatomy and indicators of function.
  5. Biological events such as brain activity involve multiple processes which all happen simultaneously. During brain activity the blood supply increases, brain cells fire off electrical impulses, chemicals: neurotransmitters are released and the whole process goes on an anatomical structure which is unique to every individual. The brain cells are bathed by a constantly changing symphony of locally released and systemic hormones. But we can only measure one change at once; measuring the electrical activity would disturb the magnetic radiation scan of the brain. The constantly changing level of transmitters is not recordable in life.  
For all these reasons, the medical imagery used in videos is based on distortions: at the macro level by the results of dissection of dead bodies, X-rays, ultrasound and at the micro level by laboratory studies of tissues and cellular processes.
My point is that none of these methods of illustration can be explanatory of daily events in the human body such as breathing, clotting, fatigue and thinking.  It's not the fault of illustrators; it's a result of the hyper-specialism of scientists who can only attribute health or disease to a factor within their specialty: a single electronic charge, a neurotransmitter, a change in structure or a pretty pattern on scans and not combinations of several of them simultaneously which our cells take for granted.

Friday 24 June 2016

What does Zika look like?

What does Zika virus look like?

 Viruses are small

Very small. Smaller than a speck of dust, smaller than a germ like E coli. Between 20 and 300nm [there are one million nanometers in a millimeter]. So we can't just take a photo of them. 
Here are some 3D renderings of viruses such as this one: Hepatitis C

It looks like a studio portrait, with two lights from above and a backlight helping it standing out from the monotone background. 

But it is nothing of the sort. It's made-up.


Viruses can only survive deep in body tissues, usually inside cells. If we could film viruses in the blood stream they would be like flotsam washing along the arteries and veins, dwarfed by thousands of giant red cells and a few even larger white cells. But we can't. Light microscopes can only reach a resolution of 200nm so even a giant virus is a minute dot.

But the electron microscope has a resolution of 50pm [there are a thousand picometers in one nanometer]. The problem is that the smallest electron microscope is about the size of a PC.  Here it is:

So it won't fit inside any living body. As a result for transmission electron microscopy you need to:

  • Take a sample
  • bring it to the lab
  • fix the structure with formaldehyde
  • freeze it quickly in liquid ethane and keep it frozen
  • take fine slices 
  • place the sample in a vacuum 
  • direct electron beams at the sample

You'll be familiar with the resulting gray scale em photos like this:


Here is our first sight of Zika virus in frozen section.

It looks spherical and transparent. There is no colour. The only way we can see coloured images is by adding stains to the sample. And the only way we can see the 3D structure of the virus is by using a scanning electron microscope, which can work with thicker sections but you may have to:
  • bathe the sample in electron dense solution
  • dry it
  • freeze fracture it to examine the broken off edge
  • embed it with epoxy resin
  • blow metal vapour such as platinum across it to create shadows
  • take an imprint of the structures with carbon vapour
  • stain it with heavy metals: lead, tungsten or uranium
As you can imagine, any of these processes may interfere with the sample to cause artefacts. So it's necessary to compare the results of several different methods to get an accurate picture.
Since viruses are essentially colourless, artificial colour is added in the staining process or in illustrating the virus structure in order to make it visible. It's guesswork:



From these sudies we know that Zika virus has a smooth surface but develops projections when it is attacked by acid inside the cell. The projections are exaggerated in some illustrations to create a more beautiful/sinister image:





 This image of Zika virus homing in on a human embryo [much reduced in size is particularly unsettling:

For me, the image which best captures the structure of Zika is this diagram: