Cortical coding of repetitive acoustic pulses

Abstract

Recordings were obtained from 179 cells, 53 of which were studied intracellularly; among these 26 cells were studied extracellularly before being studied intracellularly. A subgroup of units was characterized by spike durations of about one-third normal. These are referred to as ‘thin-spike units’. The population of cells was classified in 4 categories on the basis of responses to trains of repetitive chicks: Lockers had discharges precisely time-locked to the individual clicks. This category could be further subdivided into regular-spike lockers and thin-spike lockers. Intracellular recordings of response to single clicks from regular-spike lockers showed a short EPSP followed by a long IPSP. The limiting rate of locking for each unit was defined to be the highest rate of click repetition at which clear locking of spike activity to individual clicks in the trains could be observed. At repetition rates higher than their limiting rate regular-spike lockers usually exhibited one or two spikes at the train onset and suppression of spike activity throughout the remainder of the train. Thin-spike lockers had high rates of ongoing (spontaneous) activity and for clicks presented at low repitition rates responded with time-locked spikes and spontaneous spike activity between the locked spikes. When the limiting rate was exceeded the thin-spike lockers showed a sustained but unsynchronized discharge for the duration of the click train. Intracellular records exhibited a predominance of driven EPSPs with very little inhibitory impingement. Thirty-nine percent of the experimental population were lockers. The distribution of limiting rates was broad, from 10/sec to 1000/sec, with the median between 50/sec and 100/sec. Some lockers were studied with repetitive noise bursts. Evoked patterns of spike activity and limiting rates of these units were quite similar for clicks and brief noise pulses. The lockers exhibit a temporal coding of the repetition rate of acoustic stimuli which when presented to human listeners are judged to have periodicity pitch. Groupers had spike responses loosely synchronized to low rates of repetitive clicks. Intracellular records showed that single clicks evoked long lasting IPSPs. Spike activity was suppressed following the click sometimes returning with a rate higher than the ongoing rate. The latency of the return of activity after suppression is quite variable. For click train stimulation at low repetition rates the spike occurrences were grouped in part of the period between individual clicks. This loose synchrony was lmited to click repetition rates below 50/sec. Sixteen percent of the cells studied (but no thin-spike units) were in the grouper category. Special responders had patterns of spike activity related to the stimulus but their spikes were not synchronized with the individual clicks. The response patterns to the whole train could be described by the usual terms, on, off, through, etc. Seventeen percent of the cells studied were in this category. Twenty-eight percent of the experimental population did not respond to repetitive click stimulation. Possible neural coding mechanisms for periodicity pitch and the relationship of the present categories to those suggested on the basis of tuning curves are discussed.