Functional properties of visual tracking neurons in posterior parietal association cortex of the monkey

Abstract

The functional properties of visual tracking (VT) neurons of the posterior parietal association cortex (area 7a or PG) were studied in 10 hemispheres of 6 alert behaving [rhesus] monkeys. More strict criteria were used to classify VT neurons than the previous investigators, excluding those that responded only passively to the visual stimulation. Of 147 visual-tracking-related neurons, 70 were selected as true VT neurons, since they were activated during tracking in the dark where no visual stimulation other than the small target light was visible. The recording sites of 69 true VT neurons, determined histologically, were distributed mainly in the posterolateral part of area 7a. They were concentrated in a small region on the anterior bank of the caudal part of superior temporal sulcus (STS), near the end of lateral fissure. About 30% of them (21/69) were found in the posterior bank of STS. Only a few true VT neurons (5/69) were recorded in the anterior part of area 7a. Most of the VT neurons (54/70) had directional selectivity in the frontal plane but a few units (5/70) were multidirectional. A considerable number of units (11/70) were sensitive to the movement in depth, responding either to the approaching (9/11) or receding movement (2/11) of the target. The activity of VT neurons in the dark was compared to that in the light to assess the effect of the background image. The large proportion of them (48/70) displayed a lower discharge rate in the dark than in the light but a small proportion of units (7/70) showed a higher discharge rate in the dark. The number of VT neurons that did not show any difference was small (15/70), suggesting that most of the VT neurons are influenced by visual stimuli in the surroundings. The large majority of VT neurons tested (34/43) responded to moving visual stimulus during the fixation task. Half of them (17/34) preferred the direction of stimulus movement opposite to that of tracking and were called antidirectional VT neurons. Almost all of the antidirectional VT neurons tested showed higher activity in the lighted room where the target was tracked against stationary background. Nearly half of them (16/34) had the same directional preference of the visual stimuli as that of the tracking eye movement and were called isodirectional VT neurons. About 1/3 of the latter (6/16) increased their activity during tracking in the dark compared to that in the light, suggesting an inhibitory effect of the movement of the background image. Two of these 6 neurons actually decreased their activity when a stationary frame was present during tracking in the dark. In order to eliminate visual stimulation completely, the target light was turned off for a short period (0.5-1.0 s) during tracking. The majority of VT neurons tested (8/10) maintained a significant fraction of their activity during the period of complete darkness. VT neurons receive converging input of both the retinal signal and the extraretinal signal of eye movement. Antidirectional VT neurons are more likely to be concerned with motion perception and may be the neural basis of optical illusion, called induced movement. Isodirectional VT neurons are more likely to be concerned with the control of smooth-pursuit eye movement.