New mathematical approach could help validate leading explanation of consciousness
In a nutshell: A type of maths called information geometry takes neuroscientists one step closer to testing the Integrated Information Theory of consciousnessView Paper Abstract
Our brain constantly gathers information – thoughts, memories, sensory experiences, motivations – and encodes it in neural connections.
Consciousness is more than just the sum of this encoded information – it’s how well the pieces are melded into a whole, according to Integrated Information Theory (IIT), the current frontrunner of consciousness theories.
But is IIT really the best explanation for consciousness?
Comparing how IIT measures consciousness with alternative measures that have been proposed is colossally difficult. All use mathematical models that crunch complex brain data sets (for example, EEG readouts) in different ways. They use different measures of probability and causality going by names such as mutual information, transfer entropy, stochastic interaction and Granger causality. The aim is to put a number on consciousness states, such as coma or wide awake.
These mathematical apples and oranges can now be compared — albeit only on paper — using a form of mathematics called information geometry, according to this report by Brain Function CoE associate investigator Naotsugu Tsuchiya at Monash University, and Masafumi Oizumi and Shun-ichi Amari at the RIKEN Brain Science Institute in Japan.
The team used information geometry to compare the mathematical formulae underpinning different measures of causality used to help explain consciousness. IIT was best able to isolate causal neural interactions across the whole network rather than summing individual interactions and removing those deemed irrelevant to consciousness. Intuitively, this property makes IIT ideal to measure consciousness, say Tsuchiya.
Validating IIT is still fraught, says Tsuchiya, because the brain activity data that needs to be fed into the mathematical formulae underpinning IIT are too complex to crunch.
But the usefulness of information geometry to neuroscience extends beyond consciousness, he says. Any study where causal interactions are at play – such as where the activity of one neuron causes a neighbouring neuron to fire – could benefit from its ability to isolate meaningful interactions from background noise.
Tsuchiya’s team will measure information integration in the brain activity of people who are shown a face they consciously perceive, with those who see a quick flash of a face, which they register consciously or unconsciously. They will also attempt to determine the capacity of a simple brain, such as a fly’s, for information integration, and, potentially, consciousness.
Oizumi, M., Tsuchiya, N., & Amari, S. (2016) Unified framework for information integration based on information geometry. Proceedings of the National Academy of Sciences, 113(51), 14817-14822
Republish this article:
We believe in sharing knowledge. We use a Creative Commons Attribution 4.0 International License, which allows unrestricted use of this content, subject only to appropriate attribution. So please use this article as is, or edit it to fit your purposes. Referrals, mentions and links are appreciated.