The perturbational complexity index (PCI for short) is the outcome of applying an algorithm to electrophysiological activity evoked by a trans-cranial magnetic pulse. In brief, PCI measures how much you can "compress" the evoked activity. High compressibility implies ordered and not-very-complex activity (mostly regular), low compressibility implies a high information content.

The idea w.r.t. consciousness is that a perturbation of a system which has a small repertoire of possible states will result in a rather dull exploration of its (few) possible states, whereas a system rich with possible states will result in a more interesting and informative exploration. The first case will be associated with low complexity (low PCI), the second with high complexity (high PCI) of the evoked responses. This is the direct link with Tononi's theory, since he postulates that a high repertoire of potential possible states is crucial for consciousness, since then specifying the actual state of the system is very informative - the classical example is a photodiode with 2 possible states (light/ no light). In this case, specifying the state entails only 1 bit of information.

Besides its theoretical appeal, PCI could be practical in a clinical context since sensory stimulation (except olfaction) must pass through the thalamus to reach the cortex, and damage to the thalamus (or physiological states such as sleep, in which the thalamus becomes disengaged from the cortex) might preclude this. Direct magnetic stimulation of the cortex can overcome this issue.