RAPID COMMUNICATION Form Processing Modules in Primate Area V4

Ghose, Geoffrey M. and Daniel Y. Ts’o. Form processing modcaca fascicularis were anesthetized by continuous intravenous infusion of sufentanil citrate (3–8 mgrkgrh) , paralyzed (vercules in primate area V4. J. Neurophysiol. 77: 2191–2196, 1997. Area V4 occupies a central position among the areas of the primate uronium bromide, 100 mgrkgrh) , and artificially respirated. The level of anesthesia was continually verified by monitoring cerebral cortex involved with object recognition and analysis. Consistent with this role, neurons in V4 are selective for many visual electrocardiogram and electroencephalogram. Adequate respiration was verified by monitoring expired CO2. Craniotomies (Ç1 cm) attributes including color, orientation, and binocular disparity. However, it is uncertain whether cells within V4 are organized centered along the lunate sulcus were made to expose portions of V1, V2, and V4. The eyes were fitted with contact lenses of approwith respect to these properties. In this study we used in vivo optical imaging and electrophysiology in macaque visual cortex to priate refraction for the cathode ray tube (CRT) distance and converged by placing a Risley prism over one eye and aligning the show that cells that share certain physiological properties are indeed grouped together in V4. Our results revealed regions conmonocular receptive fields of a V2 binocular cell. Receptive fields of single neurons were mapped electrophysiologically and used to taining cells with common orientation selectivity. These regions were similar in size to those seen in V2 and much larger than those appropriately position stimuli for imaging. Images of the cortical surface were acquired by a charge-coupled device array camera seen in V1 and were confirmed by appropriately targeted singleunit recording. Surprisingly, orientation organization visible through a macroscope lens assembly during illumination by 630nm light while visual stimuli were presented in a random order on through imaging was limited to the portion of V4 representing the central visual fields. Optical imaging also revealed a functional a CRT placed 28.5 in. in front of the animal. To resolve the relatively small changes in reflectance (on the order of 1 part in organization related to stimulus size. Size-sensitive regions (S regions) contained cells that were strongly suppressed by large stim10,000) associated with neural activity, the images from 8 to 16 repetitions of each stimulus were summed. Because functional imuli. In contrast to V2, S regions in V4 contain orientation domains. These results suggest that V4 contains modular assemblies of cells ages were generated on-line, we were able to verify the increase in signal to noise as a function of repetition during the course of related to particular aspects of form analysis. Such organization may contribute to the construction of object-based representations. each experiment. Patterns of activation over the cortical surface were imaged with a spatial resolution of 43–71 mm (Ts’o et al. 1990). Functional organization with respect to a particular stimulus parameter was visualized by subtracting images acquired during I N T R O D U C T I O N the presentation of complementary visual stimuli. For example, in There has been substantial progress in our understanding Fig. 1, A and C, we constructed orientation maps by subtracting of the functional role and architecture of primary visual corimages acquired during the presentation of horizontally oriented gratings from images acquired during the presentation of vertically tex (V1), as well as several extrastriate visual areas, includoriented gratings. Stimulus sets typically included gratings of difing V2 and V5 (MT) (Albright 1984; Felleman and Van ferent spatial frequencies, spatial extents, and color modulations. Essen 1991; Hubel and Livingstone 1987; Zeki and Shipp 1988). However, despite the central position of extrastriate area V4 in the pathway devoted to object analysis (Baizer R E S U L T S et al. 1991; Tanaka et al. 1991), a definitive notion of this With the use of these methods, a functional organization area’s functional significance has not yet emerged. Although with respect to orientation was found in V4. Orientationthe initial study of V4 by Zeki (1973) suggested that the selective regions were similar in size to those found in V2 area is particularly concerned with the processing of color but considerably larger than those seen in V1 (V4: 0.15 { information, subsequent investigations have shown that V4 0.015 mm, mean { SE; V2: 0.13 { 0.015 mm; V1: cells can respond selectively to other attributes including 0.046 { 0.006 mm). The organization of these regions was orientation, binocular disparity, and size (Cheng et al. 1994; verified by recording single neurons in locations targeted Desimone et al. 1985; Desimone and Schein 1987; Gallant according to the orientation maps generated from optical et al. 1993). A similar variety of response properties is also imaging (Fig. 1A) . In six of the eight animals the stimulus present in V1 and V2 and in these visual areas neurons are properties of single neurons were examined with the use largely segregated into compartments of like functional type. of standard extracellular procedures at locations targeting To date, however, there have been no direct demonstrations according to the maps generated by imaging. A total of 76 of functional organization within V4. cells was studied from 43 penetrations. Receptive fields were qualitatively evaluated for orientation, color, and length seM E T H O D S lectivity, location, and size with the use of hand-controlled bars. For 52 of these cells, orientation selectivity was quantiTo ascertain whether such organization exists within V4, we tatively measured with the use of randomly interleaved sinuused high-resolution optical imaging of V4 along the prelunate gyrus of anesthetized and paralyzed macaque monkeys. Eight Masoidal gratings as stimuli. Orientation-selective neurons re-