Acceleration of cerebellar conditioning through improved detection of its sensory input

We studied whether slow cerebellar motor-learning is related to low signal-to-noise ratio (SNR) of the sensory signals at the pontine nucleus (PN). Rats’ PN activity was recorded during baseline and in response to auditory-stimuli. Response SNR was associated inversely with the level of baseline activity. To test the implications of this finding on cerebellar learning, a brain-computer interface was devised to analyze on-line the PN activity and deliver paired auditory conditioned stimulus (CS) and periorbital unconditioned stimulus trials, on low level baseline activity. This maneuver resulted in PN responses of high SNR and accelerated eyeblink-conditioning. These findings suggest that the poly-sensory noisy activity in the PN reduces the SNR of the CS, compromising its detection in the cerebellum which in turn slows-down the learning rate. Present results suggest a unique way in which brain-computer interfaces can be used to potentiate and expedite motor learning in healthy and brain degenerated subjects.