Last Monday dr. Aneta Stefanovska of Lancaster University Nonlinear Biomedical Physics group gave a talk at the Faculty of Mathematics and Physics in Ljubljana about their ground breaking work in modeling and interpreting signals present in humans and other living organisms. I did some research in signal analysis with her during my undergraduate study and visited the group in Lancaster almost four years ago.
I would like to share two points from the talk that I found fascinating.
They are modeling the electrical signals produced by the lower functions of the brain, heart and lungs as coupled oscillators. The idea is that each organ consists of a self-sustaining oscillator with its own resonant frequency that is to some degree affected by oscillations in other organs. This is similar to coupled oscillators in electronics (or Huygens' clocks in mechanics). The interesting point is that through signal analysis you can detect the noise entering this system of oscillators and that this noise is connected to the conscious thought processes of a person. So for example if a subject is unconscious under anesthesia the noise is gone and the oscillations are be better synchronized.
This research was mostly done for medical purposes, to develop a way of assessing the level of unconsciousness. So a part of the talk was also about how anesthetic drugs work. The leading theory is that the drugs modify the neuron's chemistry in a way that prevents rapid firing (i.e. they lengthen the time it takes to recover from an impulse). Processes required for conscious thought seem to need a higher frequency for neural connections than simple oscillators for basic life support like breathing and keeping the heart beating. So those continue to work while anything required to process data from the outside world is basically switched off.
What I understood as the message of the talk was that while we're still far away from completely understanding the human body, some things are starting to map surprisingly well to what you would expect from a mechanical or electrical machine. Imagine measuring the noise on a power rail caused by switching gates in an otherwise inaccessible VLSI CPU and trying to deduce the inner structure from it. Or for example increasing the dampening so that the gigahertz circuitry in the CPU stops working while PWM for fans and the switching power supply keeps running.