With the advent of the Integrated Information theory (IIT), we understood that the level of consciousness in a system depended on the capacity of its different elements to integrate information across multiple spatiotemporal scales, and not on the capacity of the system to transmit, encode or store information. This theory gave a clear quantifiable approach to consciousness. At this point, well-established measures were present for the latter but not for the former. Moreover, with consciousness being such a subjective experience, it was essential that its measure be reportable, robust, minimally invasive, and applicable at an individual level in a clinical setting. Quantitative theories of consciousness try to define the level and content of consciousness using mathematical structures. They can be classified under three sections:
- To model substrate systems that can be conscious
- To model the level and contents of consciousness that the substrates have
- An algorithm to equate the physical and phenomenal aspects of consciousness in a system
Behavioural (subjective) measures of consciousness started out with the most obvious approach – associating consciousness with behavioural markers or cognitive capacities, based on subjective reports. Even the inclusion of graded confidence ratings, as in the case of the perceptual awareness scale (PAS), doesn’t help to completely overcome the biases of this approach. Moreover, further research showed that this bias will always be present unless the responses are taken under extremely controlled and trained environments. Furthermore, these measures are inapplicable in cases of vegetative patients, infants, animals, deep sleep, coma and other such conditions.
Under the brain based measures of consciousness:
- Objective measures of consciousness – try to identify how sensitive a system is to the signal it receives and how it responds to it, based on signal detection theory. They try to explain how subjects’ reports are not simple, but generated by filtering internal responses to stimuli using a highly variable and context sensitive ‘criterion’. Therefore, they move beyond response bias. But, they fall short on covering the ‘integration’ aspect of consciousness while being focused on the processing of the information itself.
Therefore, the sensible direction forward for an exhaustive measure would be to combine these two approaches – have the subject report their response and then check their confidence of the report itself (known as Type 2 measure). However, a closer look makes it seem like a measure of meta-cognition or meta-awareness and not consciousness itself.
- Neurophysiological measures of consciousness – while an exciting avenue, still don’t bridge the explanatory gap in the hard problem of consciousness. Results at the neurophysiological level merely reflect those at the behavioural level.
- Informational measures of consciousness – our most viable candidates so far. Chalmers defined the two essential aspects of information in consciousness but didn’t give a measure for the same. Taking his lead, Tononi (2004) gave us IIT 1.0 which alluded to some aspects of Shannon’s Information theory and in IIT 3.0 (2014), he went further to define a whole new type of information – intrinsic information, contrary to Shannon’s concept of extrinsic information which is dependent on the presence of an external observer. Thus, IIT 3.0 doesn’t refer to information quantities such as entropy and mutual information. Seth (2005) proposed causal density (of interactions in the brain networks) as a measure of consciousness. But, overall, these measures don’t show favourable clinical viability.
Thus, in 2013, Casali presented us with Perturbational Complexity Index (PCI), a theory-driven empirical index of consciousness which reflects the information content of the brain’s response to a magnetic stimulus. It measures information integration in the thalamocortical networks in the brain in a graded manner, being able to discriminate varying levels of consciousness from states of NREM sleep, dreaming, wakefulness and sedation caused by different anaesthetic agents to states of coma (vegetative state, minimally conscious state and locked-in syndrome).
It was based on the theory that multiple brain regions, with a high degree of differentiation, integrate information to produce wakeful consciousness. Thus, brain states relating to unconsciousness or coma and related states show either very locally generated responses or stereotypical responses across a large area of the brain. Brain states for dreaming showed more complex responses of longer duration. And wakeful states showed the highest level of complexity in responses.
Its major strength lies in being independent of sensory processing and behavior. It also is flexible enough to be used for patients with structural brain injuries. Its limitation, however, is that it depicts brain states at the time of the reading and therefore, the index is not capable of predicting the fluctuations of brain states throughout the day which is the case with many patients who suffer from disorders of consciousness and also in the case of healthy individuals.
Nevertheless, it creates a solid foundation for further research into the complex spatiotemporal dynamics of the brain activity and their correlation to various consciousness levels.
MATTERS OF CONSCIOUSNESS 