A research group of Advanced Telecommunications Research Institute International (ATR), Center for Information and Neural Networks (CiNet), Cambridge University, Oxford University, and others has developed a bidirectional neurofeedback system in which humans and computer-based AI agents train each other. The result of this study shows that this interactive learning of brain-AI agents using neurofeedback can strengthen the pain control system in the human brain. If neurofeedback can be used to train people to modulate their pain, it is expected to make a significant contribution to future pain treatment, including the development of new treatments for chronic pain.
Pain control by co-adaptive learning in a brain-machine Interface
DOI : 10.1016/j.cub.2020.07.066
Suyi Zhang, Wako Yoshida, Hiroaki Mano, Takufumi Yanagisawa, Flavia Mancini, Kazuhisa Shibata, Mitsuo Kawato, Ben Seymour