Temporally Specific Sensory Signals for the Detection of Stimulus Omission in the Primate Deep Cerebellar Nuclei

3 years ago



Recordings from cerebellar dentate nucleus in monkeys trained to watch a stationary cue that appears periodically (every .3, .4, or .6 seconds, depending on which recording session), and to indicate with a saccade when either a repetition of the stimulus is skipped, or that stimulus changes shape. (1) Neurons ramp up from baseline over repeated repetitions. (2) Each repetition causes either a transient dip (90% of cells) or a transient increase (10% of cells) - but this modulation only begins after several flashes of the stimulus, as though through slow entrainment. (3) Longer periodic inter-stimulus intervals result in larger spike rate dips (4) Omission of a stimulus results in an extra ~300 ms of firing where there would normally be a dip. Transformation to an oddball stimulus causes a quicker dip. (5) Muscimol inactivation of dentate nucleus delays responses to omitted stimuli, but not to oddball stimuli. Authors conclude that dip depth is a form of sensory gain that's used for timekeeping between stimuli.




Spike rate method smoothed some events that looked (from rasters) like they should have been sharp and flat - how much "ramping" is really just, trial-to-trial jitter, smoothing, or cell-to-cell jitter (in summary plots)?

Re: Rigor