Abstract:Ideally, decisions are made based on prior knowledge, which allows for informed choices. Real life, however, often requires us to make decisions arbitrarily, without sufficient information. Decoding decision making processes from neural activity could allow for cognitive neuroprostheses and Brain-Computer Interfaces (BCIs) to support decision processes in rapid human-machine interactions, weigh decision-making confidence, and further enable neuromodulation protocols for the treatment of reward-related dysfunctions. To understand the differences between the decision-making processes in arbitrary and informed decisions, we recorded intracranial electroencephalography in a large number of cortical and subcortical areas from 5 patients during a categorization task. We demonstrate that individual decisions can be decoded from Local Field Potentials (LFPs) before motor response, in both arbitrary and informed conditions. Our analysis revealed dissimilar spatio-temporal patterns between arbitrary and informed decision-making, with arbitrary decisions being decodable in fewer brain regions and earlier in time compared to informed decisions.

Decoding Arbitrary and Informed Decisions from Intracranial Recordings in Humans