Selectivity for 3D shape that reveals distinct areas within macaque inferior temporal cortex

Science

Volumn 288, Issue 5473, 2000, Pages 2054-2056

  Janssen, Peter ^a   Vogels, Rufin ^a   Orban, Guy A ^a  

^a KU LEUVEN UNIVERSITY OF LEUVEN   (Belgium)

Author keywords

[No Author keywords available]

Indexed keywords

ARTICLE; MACACA; NEUROANATOMY; NONHUMAN; PRIORITY JOURNAL; SPATIAL SUMMATION; STRUCTURE ANALYSIS; TEMPORAL CORTEX; THREE DIMENSIONAL IMAGING;

ANIMALS; BRAIN MAPPING; ELECTRODES, IMPLANTED; FORM PERCEPTION; MACACA MULATTA; NEURONS; TEMPORAL LOBE; VISION, BINOCULAR; VISION, MONOCULAR;

ILEX COOKII; MACACA;

EID: 0034674492     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.288.5473.2054     Document Type: Article

References (28)

1. L. G. Ungerleider and M. Mishkin, in Analysis of Visual Behavior, D. J. Ingle, M. A. Goodall, R. J. Mansfield, Eds. (MIT Press, Cambridge, MA, 1982).
2. N. K. Logothetis and D. L. Sheinberg, Annu. Rev. Neurosci. 19, 577 (1996).
3. K. Tanaka, H. K. Fukuda, Y. Moriya, J. Neurophysiol. 66, 170 (1991).
4. I. Fujita, K. Tanaka, M. Ito, K. Cheng, Nature 360, 343 (1992).
5. B. Seltzer and D. N. Pandya, Brain Res. 149, 1 (1978).
6. Binocular disparity is a powerful depth cue.
7. 2 and 4 (Δ///), respectively. The maximal stimulus diameter was 5.5 degrees, dot density 50% and dot size 4 arcmin. A fixation target (24 arcmin) was superimposed on the stimulus. The fixation distance was 86 cm and the maximal amplitude of the disparity gradient was 0.65 degrees. Stimulus duration was 800 ms. Presentation of shapes filled with three different textures was inter-leaved. Two emmetropic rhesus monkeys were trained to fixate within 0.7 degree of a small target. The position of the right eye was recorded with a scleral search coil. Surgical procedures and animal treatment were in accordance with the guidelines established by the National Institutes of Health (NIH) for the care and use of laboratory animals.
8. For each vertical penetration, we noted the depths of responsive neurons, and of landmarks including white to gray matter transitions and the skull. A computer tomography (CT) scan obtained with the guiding tube in situ confirmed that the recording chambers were implanted at the targeted coordinates (16 mm anterior, 22 mm lateral). To reconstruct the recording positions, the CT and the magnetic resonance (MR) images were superimposed (Fig. 1, A and C). In Fig. 1D, all penetrations except the most lateral ones (in which no transition between STS and lateral TE was detected) were classified into three bins according to their mediolateral position, and the landmark depths were averaged for each bin separately. For the most lateral penetrations, we classified the neurons recorded in the upper 1000 μ as belonging to the lower bank of the STS, and all other neurons as lateral TE neurons. Excluding neurons recorded in this transition zone (n = 40) did not alter the results. The significance (P < 0.05) of the 3D shape-selectivity was tested by analysis of variance (ANOVA) on the net responses (interval = 400 ms). The 3D shape-selectivity was deemed not to arise from purely monocular mechanisms if the difference in response between the dichoptic presentations was at least three times greater than the difference in the sum of the monocular responses.
9. This region corresponds to TEad [K. S. Saleem and K. Tanaka, J. Neurosci. 16, 4757 (1996)].
10. ANOVA: F(3,20) = 41.3, P < 0.0001.
11. For every responsive neuron, the horizontal right eye position was analyzed in the same intervals as the neural activity. This analysis showed only small (less than 3 arcmin) eye deviations during stimulus presentation. Experiments with coils implanted in both eyes confirmed that no vergence eye movements were made upon stimulus presentation.
12. 2 = 6.5, P < 0.01).
13. In both monkeys, the total extent of cortex covered measured at least 6 mm in the anterior-posterior direction, and 5 mm in the medio-lateral direction (Fig. 1E). We made 139 penetrations (68 in lateral TE and 71 in the STS) and recorded a total of 828 neurons, 604 of which were unresponsive. Because 3D shape-selective neurons were uniformly distributed in STS, it is highly unlikely that we missed clusters of such cells in lateral TE.
14. The median 3D-shape selectivity index [(response to the preferred 3D shape) - (response to other member of the pair)/(response to preferred 3D shape)] equaled 0.82 (1st quartile: 0.66; 3rd quartile: 0.97; n = 79) for the 3D shape-selective STS neurons, compared to 0.58 for the selective neurons in the lateral TE (1st quartile: 0.35; 3rd quartile: 1.0; n = 10).
15. Non-normalized population peristimulus-time histograms (PSTHs) were computed by averaging the net activity for each population of neurons (STS, lateral TE and nonselective neurons). The population response to the nonpreferred 3D shape was subtracted binwise from the response to the preferred 3D shape. The ANOVA was computed on the resulting differential population responses. The difference between STS and lateral TE neurons proved significant [F(1,34) = 155.9, P < 0.0001].
16. i is the mean net response to each 2D shape, and max is the largest mean response. For every 2D shape, we averaged the responses to the four different 3D shapes derived from that 2D shape. The average index was nearly identical in STS (0.66) and lateral TE (0.65).
17. The proportion of 3D shape-selective neurons was very similar (10%) in a population of 76 lateral TE neurons tested with larger disparity amplitudes (1.3 degrees).
18. Kolmogorov-Smirnov two-sample test P < 0.001. For negative monocular responses, the index was set to +10 (7% of the neurons).
19. See figure 12 in J. S Baizer, L. G. Ungerleider, R. Desimone, J. Neurosci. 11, 168 (1991).
20. E. G. Jones and T. P. S. Powell, Brain 93, 793 (1970).
21. K. S. Saleem, W. Suzuki, K. Tanaka, T. Hashikawa. J. Neurosci., in press.
22. Lateral TE neurons may be sensitive to 3D structure defined by monocular depth cues, such as shading [M. Ito, I. Fujita, H. Tamura, K. Tanaka, Cerebr. Cortex 4, 499 (1994)].
23. G. Baylis, E. T. Rolls, and C. M. Leonard [J. Neurosci. 7, 330 (1987)] reported that STS neurons were more selective and more frequently responsive to faces and complex stimuli than neurons in lateral TE. Face neurons, however, can be found throughout TE [R. Desimone, J. Cogn. Neurosci. 3, 1 (1991)]. Using stimulus dimensions other than 3D shape may reveal additional differences between STS and lateral TE.
24. G. Baylis, E. T. Rolls, and C. M. Leonard [J. Neurosci. 7, 330 (1987)] reported that STS neurons were more selective and more frequently responsive to faces and complex stimuli than neurons in lateral TE. Face neurons, however, can be found throughout TE [R. Desimone, J. Cogn. Neurosci. 3, 1 (1991)]. Using stimulus dimensions other than 3D shape may reveal additional differences between STS and lateral TE.
25. E. Shikata, Y. Tanaka, H. Nakamura, M. Taira, H. Sakata, Neuroreport 7, 2389 (1996).
26. Three-dimensional information could also reach the STS through area FST [B. Seltzer and D. N. Pandya, J. Comp. Neurol. 290, 451 (1989)], which receives projections from area MT/V5 [L. G. Ungerleider and R. Desimone, J. Comp. Neurol. 248, 190 (1986)].
27. Three-dimensional information could also reach the STS through area FST [B. Seltzer and D. N. Pandya, J. Comp. Neurol. 290, 451 (1989)], which receives projections from area MT/V5 [L. G. Ungerleider and R. Desimone, J. Comp. Neurol. 248, 190 (1986)].
28. Supported by Geneeskundige Stichting Koningin Elisabeth, FWO Vlaanderen, GOA 95-99/06 and GOA 2000/11. P.J. is a research assistant and R.V. is a research associate of the FWO. We thank M. Depaep, P. Kayenbergh, G. Meulemans, G. Vanparrys, and W. Spileers.