- Email: [email protected]
Tolerances of responses to visual patterns in neurons of the posterior inferotemporal cortex in the macaque against changing stimulus size and orientation, and deleting patterns
Behavioural Brain Research 100 (1999) 67 – 76
Research report
Tolerances of responses to visual patterns in neurons of the posterior inferotemporal cortex in the macaque against changing stimulus size and orientation, and deleting patterns Kazuo Hikosaka * Department of Psychology, Tokyo Metropolitan Institute for Neuroscience, 2 – 6 Musashidai, Fuchu-city, Tokyo 183 -8526, Japan Received 11 March 1998; accepted 21 July 1998
Abstract Neuronal activities were recorded in areas TEO and TE of the inferotemporal cortex in four hemispheres of two monkeys during the performance of a visual pattern discrimination task. Tolerances of responses to patterns against changing stimulus size and orientation, and deleting patterns halves were investigated and compared between TEO and TE neurons. Of 311 neurons tested, 80 (26%) responded to one or more patterns out of four standard patterns. Of these 80 neurons, 50 (63%) were recorded in area TEO and 30 (38%) in area TE. Neurons responsive to patterns were recorded in both areas TEO and TE, however degrees of tolerance of responses were different between TEO and TE neurons. Tolerances of TEO neurons were moderate and degrees of tolerance varied from neuron to neuron. Responses to particular patterns were dependent on stimulus size, stimulus orientation, and/or completeness of patterns. By contrast, tolerances of TE neurons were generally strong. Responses to particular patterns were not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. These results suggest that area TEO rather than area TE is involved in detecting and processing particular visual shapes. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Response properties to visual patterns; Area TEO; Detecting visual shapes; Tolerances of responses; Pattern discrimination; Macaque monkey
1. Introduction The inferotemporal cortex (IT) of the macaque monkey is known to play a crucial role in the perception of visual patterns, as monkeys with bilateral damage to IT show severe deficits in visual pattern discrimination learning [3,11,22]. Based on observations of cytoarchitecture and effects of lesions, it has been proposed that IT is divided into two regions, namely area TE (anterior part of IT) and area TEO (posterior part of IT) [17,33]. Subsequent lesion studies confirmed that the * Corresponding author. Tel.: +81 42 3253881; fax: + 81 42 3218678; e-mail: [email protected]
visual function of IT is not homogenous, and area TEO rather than area TE is involved in the perception of visual patterns, whereas area TE rather than area TEO in the recognition of visual patterns [3,11,22]. Previous physiological studies have revealed that a considerable proportion of TE neurons show selectivity for a particular pattern [4,6,10,12,16,18,19,21,23,25,27 – 29,31]. Furthermore, effects of changing stimulus parameters, such as stimulus size and orientation, on responses of TE neurons have been extensively studied [16,21,23,28,29]. Invariance of response magnitude of TE neurons irrespective of changing stimulus size or orientation is considered to be important for the perceptual invariance of a particular pattern, e.g. size- and
0166-4328/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII S0166-4328(98)00114-4
68
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
orientation-constancy, that relates closely to the recognition of visual patterns [23,28]. However, it is unclear whether area TE is the first area in which neurons have tolerances of responses against changing stimulus parameters in the process of visual pattern analysis because such tolerances have not been studied in neurons of other visual areas. The information about shapes of visual patterns is analyzed in area V1 and reached to area TE via several prestriate areas and area TEO [9]. Area TEO is the last major area sending information about visual patterns to area TE [5,8,32] and contains neurons responsive selectively to particular patterns [2,6,13,18,31]. In order to examine the functional role of area TEO in the process of visual pattern analysis, activities of TEO and TE neurons were recorded during the performance of a visual pattern discrimination task. Tolerances of responses to visual patterns against changing stimulus size and orientation, and deleting patterns halves were investigated and compared between TEO and TE neurons. According to the present results, tolerances of responses of TEO neurons to visual patterns were moderate and degrees of tolerance varied from neuron to neuron. Responses were sensitive to particular patterns as well as stimulus size, stimulus orientation and/or completeness of patterns. However, tolerances of responses of TE neurons to visual patterns were generally strong. Responses were not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. These results suggest that area TEO rather than area TE is involved in detecting and processing particular visual shapes.
presentation of the ‘positive’ stimulus (GO trial). The other two patterns were ‘negative’ stimuli. The monkey was required to refrain from pressing the lever for 1 s after the onset of the presentation of the ‘negative’ stimulus (NO-GO trial). The correct responses to the negative stimuli were not rewarded. The pattern was presented for 330 ms at intervals ranging from 6 to 12 s in pseudorandom order. Before the pattern presentation, a pure tone of l kHz was given for 500 ms as a warning stimulus for the trial (Fig. 1a). Test patterns consisted of the following patterns: 1. Size-enlarging patterns. Patterns enlarged 2-fold were used. The plus and triangle patterns were presented as rewarded stimuli and the square and circle patterns, as unrewarded stimuli. 2. Size-reducing patterns. Patterns reduced 2-fold were used. The plus and triangle patterns were presented as rewarded stimuli and the square and circle patterns, as unrewarded stimuli. 3. Orientation-changing patterns. Patterns rotated in 45° from the standard plus and square patterns and the pattern rotated in 180° from the standard triangle pattern were used. The rotated plus and triangle patterns were presented as rewarded stimuli and the rotated square, as an unrewarded stimulus. 4. Patterns deleted halves. Upper and lower half of the standard plus, square, and circle patterns were used. The partial plus pattern was presented as a rewarded stimulus and the partial square and circle patterns, as unrewarded stimuli. 5. Components of patterns. Horizontal and vertical bars and bars tilted by 60 and 120° were used as unrewarded stimuli. The monkey had performed 600–1000 trials a day.
2. Materials and methods
2.1. Subjects and beha6ioral training Two male Japanese monkeys (Macaca fuscata) weighing 3.5 and 5.1 kg were used. The methods of behavioral training, physiological recording, and histological identification of recording sites were almost the same as those previously reported [13 – 15,28]. Briefly, the monkey was trained to discriminate visual patterns that were back-projected successively on the center of the translucent tangent screen subtending 30 × 30° in front of the monkey. Visual patterns consisted of the standard and of several test patterns. The standard patterns consisted of four patterns; a plus sign, an outline triangle, an outline square, and an outline circle. The size of the standard patterns was 5° with 5 cd/m2 in luminance against a background of 0.5 cd/m2. The plus and triangle patterns were ‘positive’ stimuli. Pressing a lever below the screen was rewarded with 0.5 ml of fruit juice within 500 ms after the onset of the
Fig. 1. (a) Time sequence of the visual discrimination task. The pattern stimulus was presented during 330 ms and a warning tone (1 kHz) was presented prior to the presentation of the pattern stimulus. The monkey had to press a lever after the presentation of the positive stimulus (GO trial). (b) Horizontal eye positions in ten trials during task performance. Although eye positions were relatively fixed during pattern presentation in a trial, there were some differences in eye position among trials. U, upper visual field; L, lower visual field.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
After the monkey had learned the task described above, surgery was performed under anesthesia with pentobarbital sodium (32 mg/kg). The apparatus for head fixation were permanently implanted onto the skull on the frontal and parietal lobes and two pairs of silver–silver chorides electrodes were implanted in the bones of orbital rims to obtain electro-oculography (EOG) [1]. The monkey was retrained in the task with its head fixed after recovery from surgery. Finally, a recording chamber was surgically implanted over a hole (diameter= 18 mm) in the inferotemporal bone under anesthesia. The aura was left intact and washed and cleaned daily with saline and antibiotics.
69
and orientation, and deleting patterns halves, was calculated. The tolerance ratio was defined as the mean response ratio of test patterns investigated. This analysis was done for neurons whose responses were investigated with at least two out of four kinds of test patterns such as size-enlarging patterns, size-reducing patterns, orientation-changing patterns and patterns deleted halves. The component ratio was defined as the mean response ratio of component patterns investigated. The Mann–Whitney U-test was used to analyze differences between the distributions of the tolerance and component ratios in TEO and TE neurons.
2.3. Histology 2.2. Recording procedure and analysis Extracellular recordings were made in areas TEO and TE using glass-coated platinum iridium electrodes with exposed tips of 10– 15 mm long (1 – 2 MV at l kHz). Amplified discharges were transformed to digital signals with a window discriminator and recorded on a computer disk and magnetic tape together with event signals and EOGs. Recordings were normally made during the monkey performed to discriminate the standard patterns. During the monkey performed the visual discrimination task, activities of single neurons were monitored on an oscilloscope and on a loudspeaker. The activities of neurons were calculated during two periods of 500 ms each, 500 ms prior to the presentation of the warning tone (prestimulus activity) and 50 – 550 ms after the presentation of the pattern (poststimulus activity). The latency of neurons to patterns varied between 80 and 250 ms after onset of pattern presentation and activities of most neurons were ceased until 530 ms after pattern presentation. Therefore, the specific time-locked changes in activities during 50 – 550 ms after pattern presentation were sampled and compared with prestimulus activity (Student’s t-test, two tailed). The responses of TEO and TE neurons to patterns were shown in prestimulus time histograms (PSTHs) which were compiled from 15 to 17 repetitions of each pattern. In order to investigate quantitatively the responses of IT neurons to test patterns, the response magnitudes of response-elicited patterns for each neuron were calculated. Response magnitude was defined as the mean firing rate of poststimulus activity subtracted by the mean firing rate of prestimulus activity. The response ratio of the test pattern, indicating the response magnitude to the test pattern normalized to that of the standard pattern, was calculated. In some neurons, the reduction of response magnitude to two or three kinds of test patterns was observed. To evaluate these combined effects on responses to test patterns adequately, the tolerance ratio indicating the average tolerance of responses against changing stimulus size
After the final recording session, the monkey was deeply anesthetized with pentobarbital sodium (45 mg/ kg) and perfused intracardially with warm saline followed by 10% formal saline. The brain was removed and a block of the brain was placed in the fixative containing 10% formalin and 30% sucrose until it sank. The brains were cut frozen into sections of 50 mm in the coronal plane. Every fifth section was stained for cell bodies with cresyl violet. Electrode tracks were reconstructed from traces of electrode penetrations and electrolytic lesions (10 mA, 20 s; tip negative). The boundaries between areas TEO and TE were considered to be the anterior tip of the posterior middle temporal sulcus which is well used in lesion studies [17].
3. Results Horizontal eye positions during ten trials are shown in Fig. 1b while the monkey performed the visual pattern discrimination task. Although eye positions of the monkeys were relatively fixed during pattern presentation in a trial, there were some differences in eye position among trials. Neuronal activities of TEO and TE neurons in the inferotemporal cortex (IT) were recorded during the performance of a visual pattern discrimination task in four hemispheres of two monkeys. Of 311 neurons tested for their responses to four standard patterns, 80 (26%) responded significantly to one or more patterns. Excitatory responses to visual patterns were evoked in 71 neurons (89%) and inhibitory responses were elicited in other nine neurons (11%). The recording sites of these 80 neurons are indicated by dots in the lateral views of the brains (Fig. 2). Responsive neurons were distributed widely in the middle and posterior portions of IT, extending until 20 mm anterior to the ascending limb of the inferior occipital sulcus. Based on the recording sites relative to sulci, 50 neurons (63%) were recorded in area TEO and the remaining 30 neurons (38%) in area TE.
70
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
Fig. 2. Lateral views of brains in two hemispheres of two monkeys. Dots indicate locations of neurons responsive to patterns. Abbreviations: AS, arcuate sulcus; CS, central sulcus; IOS, inferior occipital sulcus; IPS, intraparietal sulcus; LF, lateral fissure; LS, lunate sulcus; PMTS, posterior middle temporal sulcus; PS, principal sulcus; STS, superior temporal sulcus.
On the basis of responses to the standard patterns, these neurons were divided into four types; 18 neurons (Type 4) responded to all four patterns; 25 neurons (Type 3) to three patterns; 21 neurons (Type 2) to two patterns; 16 neurons (Type 1) to one particular pattern. Because neurons responsive to patterns were found in areas TEO and TE, the incidences of different response types of neurons were compared between two areas (Table 1). In areas TEO and TE, all four response types of neurons were found, and the proportion of four response types of neurons was almost the same. Furthermore, each type of neuron was distributed widely in IT, and there was no clustering of similar types of neurons observed within the recording area. These findings did not only confirm that some TE neurons responded selectively to a particular pattern during the performance of a visual pattern discrimination task, similarly to results reported previously
[28], but also showed that some TEO neurons responded selectively to a particular pattern under the same condition. However, responses to test patterns, such as size-changing patterns, orientation-changing patterns and patterns deleted halves, were different between TEO and TE neurons. Table 1 Types and population of TEO and TE neurons TEO neurons
Type Type Type Type Total
4 3 2 1
TE neurons
Exc.
Inh.
Total
Exc.
Inh.
Total
9 11 14 11
1 2 2 0
10 13 16 11
6 10 5 5
2 2 0 0
8 12 5 5
45
5
50 (100%)
26
4
30 (100%)
(20%) (26%) (32%) (22%)
(27%) (40%) (17%) (17%)
Exc. indicates excitatory responses to the standard patterns; Inh. indicates inhibitory responses to the standard patterns.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
71
Fig. 3. Responses of a TEO neuron to patterns which were affected by changing stimulus size. The pattern and the value above histograms indicate the pattern presented and the response ratio indicating the magnitude of responses to test patterns normalized to that of the standard pattern. A horizontal bar below each histogram indicates the duration of the presentation of visual patterns. The tolerance index of this TEO neuron was 0.47. Bin width, 10 ms.
3.1. Responses of TEO neurons to test patterns The response characteristics of TEO neurons to test patterns varied from neuron to neuron. Responses to patterns were affected by changing stimulus size, changing stimulus orientation and/or deleting patterns halves. One example of TEO neurons whose responses were affected by changing stimulus size is shown in Fig. 3. This neuron responded selectively to the plus pattern out of the four standard patterns. Compared to responses to the standard plus pattern, responses to the large plus pattern (response ratio= 0.57) and the small plus pattern (response ratio= 0.37) were considerably decreased. This TEO neuron was sensitive to a particular shape in a particular size. The tolerance ratio of this neuron was 0.47. Fig. 4 shows responses of a TEO neuron which were affected by changing stimulus orientation. This neuron responded selectively to the triangle out of the four standard patterns. Responses to the large triangle (response ratio= 1.09) and the small triangle (response ratio=1.1) were almost the same as those to the standard triangle. However, responses to the tri-
angle with stimulus orientation changed by 180° were decreased (response ratio= 0.19). Furthermore, this neuron was not activated by a bar tilted by 60° which is a component of the triangle (response ratio= 0.0). To activate this TEO neuron, the visual stimulus needed to be a particular shape in a particular orientation irrespective of stimulus size. The tolerance and component ratios were 0.79 and 0.0, respectively. An example of TEO neurons whose responses were affected by deleting patterns halves is shown in Fig. 5. This neuron responded selectively to the plus pattern out of the four standard patterns. Responses to the upper half of the plus pattern (response ratio= 0.55) were weaker than those to the standard plus pattern. In addition, this neuron had preference for a small rather than a larger pattern. The small plus pattern significantly activated this neuron (response ratio= 0.92), whereas the large plus pattern did not (response ratio= 0.30). To activate this neuron, the visual stimulus needed to be a particular shape in smaller stimulus size. A complete particular shape did activate this neuron, but the partial shape did not. The tolerance ratio of this neuron was 0.59.
72
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
3.2. Responses of TE neurons to test patterns Responses of TE neurons to particular patterns were generally not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. These characteristics were almost the same as those described previously [23,28]. Fig. 6 shows responses of one TE neuron to the standard and the test patterns. This neuron responded selectively to the plus pattern out of the four standard patterns. Responses to the large plus pattern (response ratio = 1.48), the small plus pattern (response ratio= 0.77), the plus pattern with stimulus orientation changed by 45° (response ratio=1.25), and the upper half of the plus pattern (response ratio= 0.85) were almost equal to those elicited by the standard plus pattern. However, responses were not elicited by a horizontal bar which is a component of the plus pattern (response ratio= 0.19). This TE neuron was sensitive to a particular shape irrespective of stimulus size and orientation. The partial pattern did activate this neuron, whereas the component of a particular pattern did not. The tolerance and component ratios were 1.09 and 0.19, respectively.
3.3. Comparison of responses to test patterns between areas TEO and TE Out of 50 TEO neurons sampled, 21 were tested with at least two out of the four kinds of test patterns (21neurons were tested with size-enlarging patterns, 21 neurons with size-reducing patterns, nine neurons with orientation-changing patterns, and four neurons with patterns deleted halves). Out of 30 TE neurons, 11 were tested with at least two out of the four kinds of test patterns (11 neurons were tested with size-enlarging patterns, 11 neurons with size-reducing patterns, four neurons with orientation-changing patterns, and two neurons with patterns deleted halves). Eight TEO neurons and five TE neurons were tested with components of patterns. TE neurons usually showed strong tolerances of responses against changing stimulus size and orientation and deleting patterns halves, whereas TEO neurons, as a whole, showed weaker tolerances against each stimulus transformation such as size increase, size decrease, orientation change and part-deletion compared with TE neurons. Such a tendency was also found in the distri-
Fig. 4. Responses of a TEO neuron to patterns which were affected by changing stimulus orientation. The tolerance and component indices of this TEO neuron were 0.79 and 0, respectively. For other conventions see Fig. 3.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
73
Fig. 5. Responses of a TEO neuron to patterns which were affected by changing stimulus size and deleting pattern halves. The tolerance index of this TEO neuron was 0.59. For other conventions see Fig. 3.
bution of the tolerance ratios indicating the average tolerance against changing stimulus size and orientation, and deleting patterns halves. Distributions of the tolerance and component ratios in areas TEO and TE are shown in Fig. 7. The distribution of the tolerance ratios ranged widely from 0.20 to 1.18 in area TEO and from 0.79 to 1.45 in area TE (mean tolerance ratio=0.77 in area TEO and 1.05 in area TE). Although distributions overlapped, the distribution in area TEO was significantly different from that in area TE (PB0.05, Mann – Whitney U-test). On the other hand, the distribution of component ratios was similar between area TEO and area TE, with the small differences being insignificant. About one third of TEO neurons (3/8) showed ratios larger than 0.25, while no TE neuron (0/5) showed ratios larger than 0.25 (mean component ratio= 0.22 in area TEO and 0.13 in area TE).
4. Discussion The finding in this study was that a considerable number of TEO and TE neurons (80/311, 26%) change
their activities during the performance of the pattern discrimination task. Therefore, these TEO and TE neurons may be involved in the analysis of visual patterns. However, degrees of tolerance of responses to visual patterns against changing stimulus size and orientation, and deleting patterns halves were different between TEO and TE neurons. These differences may reflect differences in functional roles in the perception of visual pattern between the two areas. In areas TEO and TE, neurons responding selectively to visual patterns but not to components of visual patterns were recorded, and the proportion of response types of neurons to standard patterns was almost the same. Because the analysis of visual patterns is advanced from area TEO to area TE [5,8,9,32], it would be hypothesized that the selectivity of visual pattern is still made up in area TEO. Because TE neurons had strong tolerances of responses against changing stimulus size and orientation, and deleting patterns halves, it was confirmed physiologically that area TE is involved in the recognition of visual pattern, similarly to previous findings [23,28]. Then the question arises about the functional role of area TEO in the analysis of visual patterns.
74
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
Responses of TEO neurons to particular patterns were affected by changing stimulus size, changing stimulus orientation, and/or deleting patterns halves. Tolerances of responses were moderate and degrees of tolerance varied from neuron to neuron without clustering of neurons with similar degrees of tolerance. Because eye positions were not controlled in this study, it is possible that changes in response magnitude of TEO neurons to visual patterns are due to that there were differences in eye position during pattern presentation among trials, and pattern stimuli were not always presented within the receptive fields of TEO neurons. This possibility cannot be ruled out because exact locations of receptive fields of TEO neurons were not determined. However, responses of TEO neurons to visual patterns were usually consistent among trials, although there were some differences in eye position during pattern presentation. Therefore, effects of changing stimulus parameters on response magnitude of TEO neurons would not be related to eye positions during pattern presentation among trials. Among TEO neurons with various tolerances, some neurons could detect whether the visual pattern had a
particular shape with particular size and/or orientation or not. Other neurons could discriminate whether the visual pattern was complete or not. These properties of TEO neurons would be advantageous for detecting a particular shape in particular size and orientation among various shapes in our environment that relates closely to the perception of visual patterns. In the brain, visual areas V1 and V2 consist of several subunits in which different stimulus parameters, such as orientation, color and motion, are analyzed independently [7,20,30]. Because area V2 sends projections directly to area TEO [24], it is possible that area TEO consists of several subunits in which tolerances of responses to patterns against changing particular stimulus parameters such as stimulus size and orientation are made up separately. However, neurons with various degrees of tolerance were intermingled in area TEO. Therefore, it is unlikely that tolerances develop systematically from a subunit to a subunit within area TEO. The findings of the present study also indicate that area TE is the first area in which most neurons have strong tolerances of responses to patterns against changing stimulus size, and orientation, and deleting
Fig. 6. Responses of a TE neuron to patterns which were not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. The tolerance and component indices of this TE neuron were 1.09 and 0.19, respectively. For other conventions see Fig. 3.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
75
the central visual fields to area TE are of low density [26]. Acknowledgements The author thanks Dr Takayuki Sato, Department of Behavioral Physiology, Tokyo Metropolitan Institute for Neuroscience for helpful advice throughout the course of the experiment. The author also thanks Dr Wolfram Schultz, Institute of Physiology, University of Fribourg, for correcting the English in the manuscript and Jun-ichi Hashimoto, Mitsuyo Yasuhara and Takeshi Aihra for data analysis. References
Fig. 7. Comparison of distributions of the tolerance and component ratios between areas TEO and TE. (a) The distribution of tolerance ratios. (b) The distribution of component ratios. Between areas TEO and TE, there was a significant difference in the distribution of the tolerance ratios (P B 0.05), but not that of the component ratios. Values with arrow heads indicate the mean values of the ratios in areas TEO and TE.
patterns halves, although a small proportion of TEO neurons has strong tolerances. Between areas TEO and TE neurons, degrees of tolerance were significantly different, however distributions of the tolerance ratios overlapped. Therefore, the magnitude of tolerance would develop gradually from area TEO to area TE. Such a gradual change in neuronal characteristics is also found with the selectivity for visual patterns which develops gradually from area TEO to area TE [18,31]. The present results did not reveal how the strong tolerances of responses of TE neurons are established within IT. However, one possibility would be that TE neurons with strong response tolerances against changing stimulus parameters are made up by converging information from TEO neurons sensitive to particular shapes with various levels of tolerance. This concept is also consistent with anatomical findings indicating that terminals of projections from area TEO representing
[1] Bond HW, Ho P. Solid miniature silver – silver chloride electrodes for chronic implantation. Electroencephalogr Clin Neurophysiol 1970;28:206 – 8. [2] Braitman DJ. Activity of neurons in monkey posterior temporal cortex during multidimensional visual discrimination tasks. Brain Res 1984;307:17 – 28. [3] Dean P. Effects of inferotemporal lesions on the behavior of monkeys. Psychol Bull 1976;83:41 – 71. [4] Desimone R, Albright AD, Gross CG, Bruce C. Stimulus-selective properties of inferior temporal neurons in the macaque. J Neurosci 1984;4:2051 – 62. [5] Desimone R, Fleming J, Gross CG. Prestriate afferents to inferior temporal cortex: an HRP study. Brain Res 1980;184:41–55. [6] Desimone R, Gross CG. Visual areas in the temporal cortex of the macaque. Brain Res 1979;178:363 – 80. [7] Deyoe EA, Van Essen DC. Segregation of efferent connections and receptive field properties in visual area V2 of the macaque. Nature 1985;317:58 – 61. [8] Distler C, Boussaoud D, Desimone R, Ungerleider LG. Cortical connections of inferior temporal area TEO in macaque monkeys. J Comp Neurol 1993;334:125 – 50. [9] Felleman DJ, Van Essen DC. Distributed hierarchical processing in the primate cerebral cortex. Cereb Cortex 1991;1:1 –47. [10] Fujita I, Tanaka K, Ito M, Cheng K. Columns for visual features of objects in monkey inferotemporal cortex. Nature 1992;360:343 – 6. [11] Gross CG. Visual functions of inferotemporal cortex. In: Jung R, editor. Handbook of Sensory Physiology, vol. 8 (part 3B). Berlin: Springer-Verlag, 1973:451 – 82. [12] Gross CG, Rocha-Miranda CE, Bender DB. Visual properties of neurons in inferotemporal cortex of the macaque. J Neurophysiol 1972;35:96 – 111. [13] Hikosaka K. Responsiveness of neurons in the posterior inferotemporal cortex to visual patterns in the macaque monkey. Behav Brain Res 1997;89:275 – 83. [14] Hikosaka, K. Representation of foveal visual fields in the ventral bank of the superior temporal sulcus in the posterior inferotemporal cortex of the macaque monkey. Behav Brain Res 1998 (in press). [15] Hikosaka K, Iwai E, Saito H-A, Tanaka K. Polysensory properties of neurons in the anterior bank of the caudal superior temporal sulcus of the macaque monkey. J Neurophysiol 1988;60:1615 – 37. [16] Ito M, Tamura H, Fujita I, Tanaka K. Size and position invariance of neuronal responses in monkey inferotemporal cortex. J Neurophysiol 1995;73:218 – 26.
76
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
[17] Iwai E, Mishkin M. Further evidence on the locus of the visual area in the temporal lobe of the monkey. Exp Neurol 1969;25:585 – 94. [18] Kobatake E, Tanaka K. Neuronal selectivities to complex object features in the ventral visual pathway of the macaque cerebral cortex. J Neurophysiol 1994;71:856–67. [19] Komatsu H, Ideura Y. Relationships between color, shape, and pattern selectivities of neurons in the inferior temporal cortex of the monkey. J Neurophysiol 1993;70:677–94. [20] Livingstone MS, Hubel DH. Anatomy and physiology of a color system in the primate visual cortex. J Neurosci 1984;4:309 – 56. [21] Logothetis NK, Pauls J. Psychophysical and physiological evidence for viewer-centered object representations in the primate. Cereb Cortex 1995;3:270–88. [22] Mishkin M. A memory system in monkey. Phil Trans R Soc Lond Ser B 1982;298:85 – 95. [23] Miyashita Y, Chang HS. Neuronal correlate of pictorial shortterm memory in the primate temporal cortex. Nature 1988;331:68 – 70. [24] Nakamura H, Gattass R, Desimone R, Ungerleider LG. The modular organization of projections from areas V1 and V2 to areas V4 and TEO in macaque. J Neurosci 1993;13:3681 – 91. [25] Richmond BJ, Optican LM, Podell M, Spitzer H. Temporal encoding of two dimensional patterns by single units in primate inferior temporal cortex. I. Response characteristics. J Neurophysiol 1987;57:132 – 46.
.
[26] Saleem KS, Tanaka K, Rockland KS. Specific and columnar projection from area TEO to TE in the macaque inferotemporal cortex. Cereb Cortex 1993;3:454 – 64. [27] Sary G, Vogels R, Orban GA. Cue-invariant shape selectivity of macaque inferior temporal neurons. Science 1993;260:995– 7. [28] Sato T, Kawamura T, Iwai E. Responsiveness of inferotemporal single units to visual pattern stimuli in monkeys performing discrimination. Exp Brain Res 1980;38:313 – 9. [29] Schwartz EL, Desimone R, Albright TD, Gross CG. Shape recognition and inferior temporal neurons. Proc Natl Acad Sci USA 1983;80:5776 – 8. [30] Shipp S, Zeki S. Segregation of pathways leading from area V2 to areas V4 and V5 of the macaque monkey visual cortex. Nature 1985;315:322 – 5. [31] Tanaka K, Saito H-A, Fukada Y, Moriya M. Coding visual images of objects in the inferotemporal cortex of the macaque monkey. J Neurophysiol 1991;66:170 – 89. [32] Umitsu Y, Iwai E. The posterior inferotemporal cortex in the monkey as an anatomically distinguishable area from adjacent cortical areas of the anterior inferotemporal cortex and the prestriate cortex. In: Ito M, editor. Integrative Control Functions of the Brain. Tokyo: Kodansha Scientific, 1980:384 – 6. [33] von Bonin G, Bailey P. The Neocortex of Macaca mulatta. Urbana: University of Illinois Press, 1947.
Research report
Tolerances of responses to visual patterns in neurons of the posterior inferotemporal cortex in the macaque against changing stimulus size and orientation, and deleting patterns Kazuo Hikosaka * Department of Psychology, Tokyo Metropolitan Institute for Neuroscience, 2 – 6 Musashidai, Fuchu-city, Tokyo 183 -8526, Japan Received 11 March 1998; accepted 21 July 1998
Abstract Neuronal activities were recorded in areas TEO and TE of the inferotemporal cortex in four hemispheres of two monkeys during the performance of a visual pattern discrimination task. Tolerances of responses to patterns against changing stimulus size and orientation, and deleting patterns halves were investigated and compared between TEO and TE neurons. Of 311 neurons tested, 80 (26%) responded to one or more patterns out of four standard patterns. Of these 80 neurons, 50 (63%) were recorded in area TEO and 30 (38%) in area TE. Neurons responsive to patterns were recorded in both areas TEO and TE, however degrees of tolerance of responses were different between TEO and TE neurons. Tolerances of TEO neurons were moderate and degrees of tolerance varied from neuron to neuron. Responses to particular patterns were dependent on stimulus size, stimulus orientation, and/or completeness of patterns. By contrast, tolerances of TE neurons were generally strong. Responses to particular patterns were not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. These results suggest that area TEO rather than area TE is involved in detecting and processing particular visual shapes. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Response properties to visual patterns; Area TEO; Detecting visual shapes; Tolerances of responses; Pattern discrimination; Macaque monkey
1. Introduction The inferotemporal cortex (IT) of the macaque monkey is known to play a crucial role in the perception of visual patterns, as monkeys with bilateral damage to IT show severe deficits in visual pattern discrimination learning [3,11,22]. Based on observations of cytoarchitecture and effects of lesions, it has been proposed that IT is divided into two regions, namely area TE (anterior part of IT) and area TEO (posterior part of IT) [17,33]. Subsequent lesion studies confirmed that the * Corresponding author. Tel.: +81 42 3253881; fax: + 81 42 3218678; e-mail: [email protected]
visual function of IT is not homogenous, and area TEO rather than area TE is involved in the perception of visual patterns, whereas area TE rather than area TEO in the recognition of visual patterns [3,11,22]. Previous physiological studies have revealed that a considerable proportion of TE neurons show selectivity for a particular pattern [4,6,10,12,16,18,19,21,23,25,27 – 29,31]. Furthermore, effects of changing stimulus parameters, such as stimulus size and orientation, on responses of TE neurons have been extensively studied [16,21,23,28,29]. Invariance of response magnitude of TE neurons irrespective of changing stimulus size or orientation is considered to be important for the perceptual invariance of a particular pattern, e.g. size- and
0166-4328/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII S0166-4328(98)00114-4
68
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
orientation-constancy, that relates closely to the recognition of visual patterns [23,28]. However, it is unclear whether area TE is the first area in which neurons have tolerances of responses against changing stimulus parameters in the process of visual pattern analysis because such tolerances have not been studied in neurons of other visual areas. The information about shapes of visual patterns is analyzed in area V1 and reached to area TE via several prestriate areas and area TEO [9]. Area TEO is the last major area sending information about visual patterns to area TE [5,8,32] and contains neurons responsive selectively to particular patterns [2,6,13,18,31]. In order to examine the functional role of area TEO in the process of visual pattern analysis, activities of TEO and TE neurons were recorded during the performance of a visual pattern discrimination task. Tolerances of responses to visual patterns against changing stimulus size and orientation, and deleting patterns halves were investigated and compared between TEO and TE neurons. According to the present results, tolerances of responses of TEO neurons to visual patterns were moderate and degrees of tolerance varied from neuron to neuron. Responses were sensitive to particular patterns as well as stimulus size, stimulus orientation and/or completeness of patterns. However, tolerances of responses of TE neurons to visual patterns were generally strong. Responses were not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. These results suggest that area TEO rather than area TE is involved in detecting and processing particular visual shapes.
presentation of the ‘positive’ stimulus (GO trial). The other two patterns were ‘negative’ stimuli. The monkey was required to refrain from pressing the lever for 1 s after the onset of the presentation of the ‘negative’ stimulus (NO-GO trial). The correct responses to the negative stimuli were not rewarded. The pattern was presented for 330 ms at intervals ranging from 6 to 12 s in pseudorandom order. Before the pattern presentation, a pure tone of l kHz was given for 500 ms as a warning stimulus for the trial (Fig. 1a). Test patterns consisted of the following patterns: 1. Size-enlarging patterns. Patterns enlarged 2-fold were used. The plus and triangle patterns were presented as rewarded stimuli and the square and circle patterns, as unrewarded stimuli. 2. Size-reducing patterns. Patterns reduced 2-fold were used. The plus and triangle patterns were presented as rewarded stimuli and the square and circle patterns, as unrewarded stimuli. 3. Orientation-changing patterns. Patterns rotated in 45° from the standard plus and square patterns and the pattern rotated in 180° from the standard triangle pattern were used. The rotated plus and triangle patterns were presented as rewarded stimuli and the rotated square, as an unrewarded stimulus. 4. Patterns deleted halves. Upper and lower half of the standard plus, square, and circle patterns were used. The partial plus pattern was presented as a rewarded stimulus and the partial square and circle patterns, as unrewarded stimuli. 5. Components of patterns. Horizontal and vertical bars and bars tilted by 60 and 120° were used as unrewarded stimuli. The monkey had performed 600–1000 trials a day.
2. Materials and methods
2.1. Subjects and beha6ioral training Two male Japanese monkeys (Macaca fuscata) weighing 3.5 and 5.1 kg were used. The methods of behavioral training, physiological recording, and histological identification of recording sites were almost the same as those previously reported [13 – 15,28]. Briefly, the monkey was trained to discriminate visual patterns that were back-projected successively on the center of the translucent tangent screen subtending 30 × 30° in front of the monkey. Visual patterns consisted of the standard and of several test patterns. The standard patterns consisted of four patterns; a plus sign, an outline triangle, an outline square, and an outline circle. The size of the standard patterns was 5° with 5 cd/m2 in luminance against a background of 0.5 cd/m2. The plus and triangle patterns were ‘positive’ stimuli. Pressing a lever below the screen was rewarded with 0.5 ml of fruit juice within 500 ms after the onset of the
Fig. 1. (a) Time sequence of the visual discrimination task. The pattern stimulus was presented during 330 ms and a warning tone (1 kHz) was presented prior to the presentation of the pattern stimulus. The monkey had to press a lever after the presentation of the positive stimulus (GO trial). (b) Horizontal eye positions in ten trials during task performance. Although eye positions were relatively fixed during pattern presentation in a trial, there were some differences in eye position among trials. U, upper visual field; L, lower visual field.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
After the monkey had learned the task described above, surgery was performed under anesthesia with pentobarbital sodium (32 mg/kg). The apparatus for head fixation were permanently implanted onto the skull on the frontal and parietal lobes and two pairs of silver–silver chorides electrodes were implanted in the bones of orbital rims to obtain electro-oculography (EOG) [1]. The monkey was retrained in the task with its head fixed after recovery from surgery. Finally, a recording chamber was surgically implanted over a hole (diameter= 18 mm) in the inferotemporal bone under anesthesia. The aura was left intact and washed and cleaned daily with saline and antibiotics.
69
and orientation, and deleting patterns halves, was calculated. The tolerance ratio was defined as the mean response ratio of test patterns investigated. This analysis was done for neurons whose responses were investigated with at least two out of four kinds of test patterns such as size-enlarging patterns, size-reducing patterns, orientation-changing patterns and patterns deleted halves. The component ratio was defined as the mean response ratio of component patterns investigated. The Mann–Whitney U-test was used to analyze differences between the distributions of the tolerance and component ratios in TEO and TE neurons.
2.3. Histology 2.2. Recording procedure and analysis Extracellular recordings were made in areas TEO and TE using glass-coated platinum iridium electrodes with exposed tips of 10– 15 mm long (1 – 2 MV at l kHz). Amplified discharges were transformed to digital signals with a window discriminator and recorded on a computer disk and magnetic tape together with event signals and EOGs. Recordings were normally made during the monkey performed to discriminate the standard patterns. During the monkey performed the visual discrimination task, activities of single neurons were monitored on an oscilloscope and on a loudspeaker. The activities of neurons were calculated during two periods of 500 ms each, 500 ms prior to the presentation of the warning tone (prestimulus activity) and 50 – 550 ms after the presentation of the pattern (poststimulus activity). The latency of neurons to patterns varied between 80 and 250 ms after onset of pattern presentation and activities of most neurons were ceased until 530 ms after pattern presentation. Therefore, the specific time-locked changes in activities during 50 – 550 ms after pattern presentation were sampled and compared with prestimulus activity (Student’s t-test, two tailed). The responses of TEO and TE neurons to patterns were shown in prestimulus time histograms (PSTHs) which were compiled from 15 to 17 repetitions of each pattern. In order to investigate quantitatively the responses of IT neurons to test patterns, the response magnitudes of response-elicited patterns for each neuron were calculated. Response magnitude was defined as the mean firing rate of poststimulus activity subtracted by the mean firing rate of prestimulus activity. The response ratio of the test pattern, indicating the response magnitude to the test pattern normalized to that of the standard pattern, was calculated. In some neurons, the reduction of response magnitude to two or three kinds of test patterns was observed. To evaluate these combined effects on responses to test patterns adequately, the tolerance ratio indicating the average tolerance of responses against changing stimulus size
After the final recording session, the monkey was deeply anesthetized with pentobarbital sodium (45 mg/ kg) and perfused intracardially with warm saline followed by 10% formal saline. The brain was removed and a block of the brain was placed in the fixative containing 10% formalin and 30% sucrose until it sank. The brains were cut frozen into sections of 50 mm in the coronal plane. Every fifth section was stained for cell bodies with cresyl violet. Electrode tracks were reconstructed from traces of electrode penetrations and electrolytic lesions (10 mA, 20 s; tip negative). The boundaries between areas TEO and TE were considered to be the anterior tip of the posterior middle temporal sulcus which is well used in lesion studies [17].
3. Results Horizontal eye positions during ten trials are shown in Fig. 1b while the monkey performed the visual pattern discrimination task. Although eye positions of the monkeys were relatively fixed during pattern presentation in a trial, there were some differences in eye position among trials. Neuronal activities of TEO and TE neurons in the inferotemporal cortex (IT) were recorded during the performance of a visual pattern discrimination task in four hemispheres of two monkeys. Of 311 neurons tested for their responses to four standard patterns, 80 (26%) responded significantly to one or more patterns. Excitatory responses to visual patterns were evoked in 71 neurons (89%) and inhibitory responses were elicited in other nine neurons (11%). The recording sites of these 80 neurons are indicated by dots in the lateral views of the brains (Fig. 2). Responsive neurons were distributed widely in the middle and posterior portions of IT, extending until 20 mm anterior to the ascending limb of the inferior occipital sulcus. Based on the recording sites relative to sulci, 50 neurons (63%) were recorded in area TEO and the remaining 30 neurons (38%) in area TE.
70
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
Fig. 2. Lateral views of brains in two hemispheres of two monkeys. Dots indicate locations of neurons responsive to patterns. Abbreviations: AS, arcuate sulcus; CS, central sulcus; IOS, inferior occipital sulcus; IPS, intraparietal sulcus; LF, lateral fissure; LS, lunate sulcus; PMTS, posterior middle temporal sulcus; PS, principal sulcus; STS, superior temporal sulcus.
On the basis of responses to the standard patterns, these neurons were divided into four types; 18 neurons (Type 4) responded to all four patterns; 25 neurons (Type 3) to three patterns; 21 neurons (Type 2) to two patterns; 16 neurons (Type 1) to one particular pattern. Because neurons responsive to patterns were found in areas TEO and TE, the incidences of different response types of neurons were compared between two areas (Table 1). In areas TEO and TE, all four response types of neurons were found, and the proportion of four response types of neurons was almost the same. Furthermore, each type of neuron was distributed widely in IT, and there was no clustering of similar types of neurons observed within the recording area. These findings did not only confirm that some TE neurons responded selectively to a particular pattern during the performance of a visual pattern discrimination task, similarly to results reported previously
[28], but also showed that some TEO neurons responded selectively to a particular pattern under the same condition. However, responses to test patterns, such as size-changing patterns, orientation-changing patterns and patterns deleted halves, were different between TEO and TE neurons. Table 1 Types and population of TEO and TE neurons TEO neurons
Type Type Type Type Total
4 3 2 1
TE neurons
Exc.
Inh.
Total
Exc.
Inh.
Total
9 11 14 11
1 2 2 0
10 13 16 11
6 10 5 5
2 2 0 0
8 12 5 5
45
5
50 (100%)
26
4
30 (100%)
(20%) (26%) (32%) (22%)
(27%) (40%) (17%) (17%)
Exc. indicates excitatory responses to the standard patterns; Inh. indicates inhibitory responses to the standard patterns.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
71
Fig. 3. Responses of a TEO neuron to patterns which were affected by changing stimulus size. The pattern and the value above histograms indicate the pattern presented and the response ratio indicating the magnitude of responses to test patterns normalized to that of the standard pattern. A horizontal bar below each histogram indicates the duration of the presentation of visual patterns. The tolerance index of this TEO neuron was 0.47. Bin width, 10 ms.
3.1. Responses of TEO neurons to test patterns The response characteristics of TEO neurons to test patterns varied from neuron to neuron. Responses to patterns were affected by changing stimulus size, changing stimulus orientation and/or deleting patterns halves. One example of TEO neurons whose responses were affected by changing stimulus size is shown in Fig. 3. This neuron responded selectively to the plus pattern out of the four standard patterns. Compared to responses to the standard plus pattern, responses to the large plus pattern (response ratio= 0.57) and the small plus pattern (response ratio= 0.37) were considerably decreased. This TEO neuron was sensitive to a particular shape in a particular size. The tolerance ratio of this neuron was 0.47. Fig. 4 shows responses of a TEO neuron which were affected by changing stimulus orientation. This neuron responded selectively to the triangle out of the four standard patterns. Responses to the large triangle (response ratio= 1.09) and the small triangle (response ratio=1.1) were almost the same as those to the standard triangle. However, responses to the tri-
angle with stimulus orientation changed by 180° were decreased (response ratio= 0.19). Furthermore, this neuron was not activated by a bar tilted by 60° which is a component of the triangle (response ratio= 0.0). To activate this TEO neuron, the visual stimulus needed to be a particular shape in a particular orientation irrespective of stimulus size. The tolerance and component ratios were 0.79 and 0.0, respectively. An example of TEO neurons whose responses were affected by deleting patterns halves is shown in Fig. 5. This neuron responded selectively to the plus pattern out of the four standard patterns. Responses to the upper half of the plus pattern (response ratio= 0.55) were weaker than those to the standard plus pattern. In addition, this neuron had preference for a small rather than a larger pattern. The small plus pattern significantly activated this neuron (response ratio= 0.92), whereas the large plus pattern did not (response ratio= 0.30). To activate this neuron, the visual stimulus needed to be a particular shape in smaller stimulus size. A complete particular shape did activate this neuron, but the partial shape did not. The tolerance ratio of this neuron was 0.59.
72
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
3.2. Responses of TE neurons to test patterns Responses of TE neurons to particular patterns were generally not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. These characteristics were almost the same as those described previously [23,28]. Fig. 6 shows responses of one TE neuron to the standard and the test patterns. This neuron responded selectively to the plus pattern out of the four standard patterns. Responses to the large plus pattern (response ratio = 1.48), the small plus pattern (response ratio= 0.77), the plus pattern with stimulus orientation changed by 45° (response ratio=1.25), and the upper half of the plus pattern (response ratio= 0.85) were almost equal to those elicited by the standard plus pattern. However, responses were not elicited by a horizontal bar which is a component of the plus pattern (response ratio= 0.19). This TE neuron was sensitive to a particular shape irrespective of stimulus size and orientation. The partial pattern did activate this neuron, whereas the component of a particular pattern did not. The tolerance and component ratios were 1.09 and 0.19, respectively.
3.3. Comparison of responses to test patterns between areas TEO and TE Out of 50 TEO neurons sampled, 21 were tested with at least two out of the four kinds of test patterns (21neurons were tested with size-enlarging patterns, 21 neurons with size-reducing patterns, nine neurons with orientation-changing patterns, and four neurons with patterns deleted halves). Out of 30 TE neurons, 11 were tested with at least two out of the four kinds of test patterns (11 neurons were tested with size-enlarging patterns, 11 neurons with size-reducing patterns, four neurons with orientation-changing patterns, and two neurons with patterns deleted halves). Eight TEO neurons and five TE neurons were tested with components of patterns. TE neurons usually showed strong tolerances of responses against changing stimulus size and orientation and deleting patterns halves, whereas TEO neurons, as a whole, showed weaker tolerances against each stimulus transformation such as size increase, size decrease, orientation change and part-deletion compared with TE neurons. Such a tendency was also found in the distri-
Fig. 4. Responses of a TEO neuron to patterns which were affected by changing stimulus orientation. The tolerance and component indices of this TEO neuron were 0.79 and 0, respectively. For other conventions see Fig. 3.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
73
Fig. 5. Responses of a TEO neuron to patterns which were affected by changing stimulus size and deleting pattern halves. The tolerance index of this TEO neuron was 0.59. For other conventions see Fig. 3.
bution of the tolerance ratios indicating the average tolerance against changing stimulus size and orientation, and deleting patterns halves. Distributions of the tolerance and component ratios in areas TEO and TE are shown in Fig. 7. The distribution of the tolerance ratios ranged widely from 0.20 to 1.18 in area TEO and from 0.79 to 1.45 in area TE (mean tolerance ratio=0.77 in area TEO and 1.05 in area TE). Although distributions overlapped, the distribution in area TEO was significantly different from that in area TE (PB0.05, Mann – Whitney U-test). On the other hand, the distribution of component ratios was similar between area TEO and area TE, with the small differences being insignificant. About one third of TEO neurons (3/8) showed ratios larger than 0.25, while no TE neuron (0/5) showed ratios larger than 0.25 (mean component ratio= 0.22 in area TEO and 0.13 in area TE).
4. Discussion The finding in this study was that a considerable number of TEO and TE neurons (80/311, 26%) change
their activities during the performance of the pattern discrimination task. Therefore, these TEO and TE neurons may be involved in the analysis of visual patterns. However, degrees of tolerance of responses to visual patterns against changing stimulus size and orientation, and deleting patterns halves were different between TEO and TE neurons. These differences may reflect differences in functional roles in the perception of visual pattern between the two areas. In areas TEO and TE, neurons responding selectively to visual patterns but not to components of visual patterns were recorded, and the proportion of response types of neurons to standard patterns was almost the same. Because the analysis of visual patterns is advanced from area TEO to area TE [5,8,9,32], it would be hypothesized that the selectivity of visual pattern is still made up in area TEO. Because TE neurons had strong tolerances of responses against changing stimulus size and orientation, and deleting patterns halves, it was confirmed physiologically that area TE is involved in the recognition of visual pattern, similarly to previous findings [23,28]. Then the question arises about the functional role of area TEO in the analysis of visual patterns.
74
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
Responses of TEO neurons to particular patterns were affected by changing stimulus size, changing stimulus orientation, and/or deleting patterns halves. Tolerances of responses were moderate and degrees of tolerance varied from neuron to neuron without clustering of neurons with similar degrees of tolerance. Because eye positions were not controlled in this study, it is possible that changes in response magnitude of TEO neurons to visual patterns are due to that there were differences in eye position during pattern presentation among trials, and pattern stimuli were not always presented within the receptive fields of TEO neurons. This possibility cannot be ruled out because exact locations of receptive fields of TEO neurons were not determined. However, responses of TEO neurons to visual patterns were usually consistent among trials, although there were some differences in eye position during pattern presentation. Therefore, effects of changing stimulus parameters on response magnitude of TEO neurons would not be related to eye positions during pattern presentation among trials. Among TEO neurons with various tolerances, some neurons could detect whether the visual pattern had a
particular shape with particular size and/or orientation or not. Other neurons could discriminate whether the visual pattern was complete or not. These properties of TEO neurons would be advantageous for detecting a particular shape in particular size and orientation among various shapes in our environment that relates closely to the perception of visual patterns. In the brain, visual areas V1 and V2 consist of several subunits in which different stimulus parameters, such as orientation, color and motion, are analyzed independently [7,20,30]. Because area V2 sends projections directly to area TEO [24], it is possible that area TEO consists of several subunits in which tolerances of responses to patterns against changing particular stimulus parameters such as stimulus size and orientation are made up separately. However, neurons with various degrees of tolerance were intermingled in area TEO. Therefore, it is unlikely that tolerances develop systematically from a subunit to a subunit within area TEO. The findings of the present study also indicate that area TE is the first area in which most neurons have strong tolerances of responses to patterns against changing stimulus size, and orientation, and deleting
Fig. 6. Responses of a TE neuron to patterns which were not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. The tolerance and component indices of this TE neuron were 1.09 and 0.19, respectively. For other conventions see Fig. 3.
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
75
the central visual fields to area TE are of low density [26]. Acknowledgements The author thanks Dr Takayuki Sato, Department of Behavioral Physiology, Tokyo Metropolitan Institute for Neuroscience for helpful advice throughout the course of the experiment. The author also thanks Dr Wolfram Schultz, Institute of Physiology, University of Fribourg, for correcting the English in the manuscript and Jun-ichi Hashimoto, Mitsuyo Yasuhara and Takeshi Aihra for data analysis. References
Fig. 7. Comparison of distributions of the tolerance and component ratios between areas TEO and TE. (a) The distribution of tolerance ratios. (b) The distribution of component ratios. Between areas TEO and TE, there was a significant difference in the distribution of the tolerance ratios (P B 0.05), but not that of the component ratios. Values with arrow heads indicate the mean values of the ratios in areas TEO and TE.
patterns halves, although a small proportion of TEO neurons has strong tolerances. Between areas TEO and TE neurons, degrees of tolerance were significantly different, however distributions of the tolerance ratios overlapped. Therefore, the magnitude of tolerance would develop gradually from area TEO to area TE. Such a gradual change in neuronal characteristics is also found with the selectivity for visual patterns which develops gradually from area TEO to area TE [18,31]. The present results did not reveal how the strong tolerances of responses of TE neurons are established within IT. However, one possibility would be that TE neurons with strong response tolerances against changing stimulus parameters are made up by converging information from TEO neurons sensitive to particular shapes with various levels of tolerance. This concept is also consistent with anatomical findings indicating that terminals of projections from area TEO representing
[1] Bond HW, Ho P. Solid miniature silver – silver chloride electrodes for chronic implantation. Electroencephalogr Clin Neurophysiol 1970;28:206 – 8. [2] Braitman DJ. Activity of neurons in monkey posterior temporal cortex during multidimensional visual discrimination tasks. Brain Res 1984;307:17 – 28. [3] Dean P. Effects of inferotemporal lesions on the behavior of monkeys. Psychol Bull 1976;83:41 – 71. [4] Desimone R, Albright AD, Gross CG, Bruce C. Stimulus-selective properties of inferior temporal neurons in the macaque. J Neurosci 1984;4:2051 – 62. [5] Desimone R, Fleming J, Gross CG. Prestriate afferents to inferior temporal cortex: an HRP study. Brain Res 1980;184:41–55. [6] Desimone R, Gross CG. Visual areas in the temporal cortex of the macaque. Brain Res 1979;178:363 – 80. [7] Deyoe EA, Van Essen DC. Segregation of efferent connections and receptive field properties in visual area V2 of the macaque. Nature 1985;317:58 – 61. [8] Distler C, Boussaoud D, Desimone R, Ungerleider LG. Cortical connections of inferior temporal area TEO in macaque monkeys. J Comp Neurol 1993;334:125 – 50. [9] Felleman DJ, Van Essen DC. Distributed hierarchical processing in the primate cerebral cortex. Cereb Cortex 1991;1:1 –47. [10] Fujita I, Tanaka K, Ito M, Cheng K. Columns for visual features of objects in monkey inferotemporal cortex. Nature 1992;360:343 – 6. [11] Gross CG. Visual functions of inferotemporal cortex. In: Jung R, editor. Handbook of Sensory Physiology, vol. 8 (part 3B). Berlin: Springer-Verlag, 1973:451 – 82. [12] Gross CG, Rocha-Miranda CE, Bender DB. Visual properties of neurons in inferotemporal cortex of the macaque. J Neurophysiol 1972;35:96 – 111. [13] Hikosaka K. Responsiveness of neurons in the posterior inferotemporal cortex to visual patterns in the macaque monkey. Behav Brain Res 1997;89:275 – 83. [14] Hikosaka, K. Representation of foveal visual fields in the ventral bank of the superior temporal sulcus in the posterior inferotemporal cortex of the macaque monkey. Behav Brain Res 1998 (in press). [15] Hikosaka K, Iwai E, Saito H-A, Tanaka K. Polysensory properties of neurons in the anterior bank of the caudal superior temporal sulcus of the macaque monkey. J Neurophysiol 1988;60:1615 – 37. [16] Ito M, Tamura H, Fujita I, Tanaka K. Size and position invariance of neuronal responses in monkey inferotemporal cortex. J Neurophysiol 1995;73:218 – 26.
76
K. Hikosaka / Beha6ioural Brain Research 100 (1999) 67–76
[17] Iwai E, Mishkin M. Further evidence on the locus of the visual area in the temporal lobe of the monkey. Exp Neurol 1969;25:585 – 94. [18] Kobatake E, Tanaka K. Neuronal selectivities to complex object features in the ventral visual pathway of the macaque cerebral cortex. J Neurophysiol 1994;71:856–67. [19] Komatsu H, Ideura Y. Relationships between color, shape, and pattern selectivities of neurons in the inferior temporal cortex of the monkey. J Neurophysiol 1993;70:677–94. [20] Livingstone MS, Hubel DH. Anatomy and physiology of a color system in the primate visual cortex. J Neurosci 1984;4:309 – 56. [21] Logothetis NK, Pauls J. Psychophysical and physiological evidence for viewer-centered object representations in the primate. Cereb Cortex 1995;3:270–88. [22] Mishkin M. A memory system in monkey. Phil Trans R Soc Lond Ser B 1982;298:85 – 95. [23] Miyashita Y, Chang HS. Neuronal correlate of pictorial shortterm memory in the primate temporal cortex. Nature 1988;331:68 – 70. [24] Nakamura H, Gattass R, Desimone R, Ungerleider LG. The modular organization of projections from areas V1 and V2 to areas V4 and TEO in macaque. J Neurosci 1993;13:3681 – 91. [25] Richmond BJ, Optican LM, Podell M, Spitzer H. Temporal encoding of two dimensional patterns by single units in primate inferior temporal cortex. I. Response characteristics. J Neurophysiol 1987;57:132 – 46.
.
[26] Saleem KS, Tanaka K, Rockland KS. Specific and columnar projection from area TEO to TE in the macaque inferotemporal cortex. Cereb Cortex 1993;3:454 – 64. [27] Sary G, Vogels R, Orban GA. Cue-invariant shape selectivity of macaque inferior temporal neurons. Science 1993;260:995– 7. [28] Sato T, Kawamura T, Iwai E. Responsiveness of inferotemporal single units to visual pattern stimuli in monkeys performing discrimination. Exp Brain Res 1980;38:313 – 9. [29] Schwartz EL, Desimone R, Albright TD, Gross CG. Shape recognition and inferior temporal neurons. Proc Natl Acad Sci USA 1983;80:5776 – 8. [30] Shipp S, Zeki S. Segregation of pathways leading from area V2 to areas V4 and V5 of the macaque monkey visual cortex. Nature 1985;315:322 – 5. [31] Tanaka K, Saito H-A, Fukada Y, Moriya M. Coding visual images of objects in the inferotemporal cortex of the macaque monkey. J Neurophysiol 1991;66:170 – 89. [32] Umitsu Y, Iwai E. The posterior inferotemporal cortex in the monkey as an anatomically distinguishable area from adjacent cortical areas of the anterior inferotemporal cortex and the prestriate cortex. In: Ito M, editor. Integrative Control Functions of the Brain. Tokyo: Kodansha Scientific, 1980:384 – 6. [33] von Bonin G, Bailey P. The Neocortex of Macaca mulatta. Urbana: University of Illinois Press, 1947.