The Primate Visual System PDF

Preface

The chapters in this book focus on the primate visual system, but not the entire primate visual system. The emphasis is on parts of the system that seem specialized in primates and thus notably different from those in other taxa of mammals. This specialization starts in the retina, where major modifications in the proportion of ganglion cell classes, their distribution across the retina, and their central targets are essential for the emphasis in primates on detailed frontal vision (Silveira, Chapter 2; Casagrande and Royal, Chapter 1). In particular, vast numbers of ganglion cells are concentrated in the central retina, and most of these cells (roughly 80%) are of the parvocellular (P cell) class, which mediates form vision and object recognition. Basically, all of these P cells, together with the magnocellular (M cell) and koniocellular (K cell) classes, project to the dorsal lateral geniculate nucleus (LGN) of the visual thalamus. The LGN has a characteristic pattern of lamination, based on the segregation of inputs according to ganglion cell class and eye of origin, so that an experienced investigator can easily identify the LGN as from a primate by appearance alone. In addition, modifications of the overall pattern of lamination allow one to identify the structure as belonging to one of the three major branches of the primate radiation (prosimian primates, tarsiers, and anthropoid primates). In diurnal anthropoid primates (monkeys, apes, and humans), a great expansion of the P cell layers, reflecting the importance of detailed central vision and color vision, is obvious.

Just medial and caudal to the LGN, the pulvinar complex of the visual thalamus is also greatly expanded compared with the pulvinar (often termed the lateral posterior nucleus or complex) of other mammals. Parts of the pulvinar complex receive visual inputs from the superior colliculus, and some visual inputs come directly from the retina, but the uniqueness of the pulvinar complex in primates arises from its relationship with the visual cortex. Because most of the inputs of the visual pulvinar are from subdivisions of visual cortex and because all of the outputs are to subdivisions of visual cortex, the complexity and size of the visual pulvinar are closely tied to the great expansion and complexity of the visual cortex in primates, especially in the anthropoid primates. Recently, there has been much progress in determining how the pulvinar complex is divided into nuclei, and how these nuclei are interconnected with areas of visual cortex. These recent findings are nicely reviewed by Stepniewska (Chapter 3).

In primates, nearly all projections of the LGN are to primary visual cortex (V1), and this area is responsible for directly or indirectly activating most of the rest of visual cortex. Although much could be said about the many studies of the internal organization of V1, the response properties of neurons in V1, and the roles of local circuits in producing these response properties, such extensive coverage of even a major visual area is not possible in this book. Instead, we concentrate on the very recent progress in our understanding of how the neurons in V1 of monkeys emerge in development (Chino et al., Chapter 4). Surprisingly, neurons have many adultlike properties soon after birth, suggesting that the development of these properties does not depend on postnatal visual experience. In addition, inputs, outputs, internal organization, and other features of V1 are covered in other chapters (Casagrande and Royal, Chapter 1; Stepniewska, Chapter 3; Roe, Chapter 5; Kaas, Chapter 6; Bullier, Chapter 8; Preuss, Chapter 10; Rosa and Tweedale, Chapter 11).

In proportional size, the second visual area, V2, is the next largest (after V1) visual area, and great progress has been made in understanding the nature of the modular organization of V2, how this organization relates to the processing of types of visual information, and how these modules receive different distributions of inputs from V1 and project to other visual areas, especially DL (V4) and MT (V5). Anna Roe's studies (Chapter 5) have been in the forefront of those producing this progress.

Primates clearly have a large number of cortical areas that are involved in processing visual information. Some investigators estimate that macaque monkeys have as many as 35 areas that are predominantly visual in function. This number is likely to vary across primate taxa, with prosimian primates having fewer visual areas and humans having more, but most visual areas have not been well defined (see Kaas, Chapter 6), and the exact number of visual areas is not known for any primate. However, it is generally recognized that the visual areas can be assigned to levels in a processing hierarchy, with V1 in the early distribution center, V2 as the major cortical target of V1 at the second level, areas with direct inputs from V1 and V2 at a third level, and so on. The problem is that the large number of visual areas and the extensive connection framework they produce are so complex that different hierarchies can be constructed from the data, depending on the assumptions of the investigator and the data considered. As such hierarchies provide a useful framework for considering how visual cortex mediates functions, a chapter is specifically devoted to discussing visual hierarchies, including parallel components (Bullier, Chapter 8; also see Casagrande and Royal, Chapter 1). Areas early in the hierarchy are considered in one chapter (Kaas, Chapter 6), these and other areas are described further in a chapter especially devoted to the mapping features of visual cortex (Rosa and Tweedale, Chapter 11), and higher level areas of inferotemporal cortex of the ventral stream of processing for object vision are described in a third chapter (Tanifuji, Chapter 14). The classical visuomotor stream of processing involves subdivisions of posterior parietal cortex and visuomotor areas of the frontal lobe (Schall et al., Chapter 9). Thus, the major organizational features of visual cortex in primates are covered by a collection of chapters by leading investigators.

Other chapters are devoted to specific features of the visual cortex. In general, investigators have been concerned with the driving, feedforward connections in the visual cortex, but visual areas also project back to the areas that provide their inputs. Overall, there has been growing interest in determining what these feedback connections do. Fortunately, we have an excellent chapter (Rockland, Chapter 16) by a pioneer who first described the major anatomical features of cortical feedback connections. A major cell type in cortex is the pyramidal cell, and it came as a great surprise to many that the dendritic arbors and other aspects of pyramidal cell morphology vary across visual areas and across primate species. As such variations in cell morphology reflect specialization for different functional roles, studies of pyramidal cells in visual cortex provide uniquely new insights into cortical processing (Elston, Chapter 15). The related chapter by Preuss (Chapter 10) addresses the issue of specialization in the retina, LGN, and areas of visual cortex. More specifically, Preuss demonstrates that there are marked differences in the internal organizations of these structures in humans and monkeys, clearly suggesting that the visual systems of these primates do not function in exactly the same ways.

Another issue that has drawn the attention of researchers recently is the plasticity of the mature visual system. Lesions of the retina and cortex are followed by compensations in the visual system that we are just now beginning to describe and understand (Collins and Kaas, Chapter 7). Such recoveries may relate to the perceptual process of "filling-in" of blind spots in the visual field, and the preservation and recovery of visual abilities, possibly including some of those described as "blindsight."

Finally, great progress has been made in understanding the organization and functions of the human visual system. Preuss (Chapter 10) describes some of the anatomical specializations of the human visual system, and Blake et al. (Chapter 13) outline how visual motion is processed in human visual cortex. Humans are especially good at identifying faces, and a region of the human temporal lobe called the "fusiform face area" has been identified as an area involved in this ability. Gauthier addresses the intriguing issue of what learning about face processing in the visual cortex tells us about how the visual system mediates object recognition (Chapter 12).

Overall, these chapters by leading investigators provide a current and extensive review of the neuronal mechanisms of visual perception and action in the visual system of primates. Interesting species differences and specializations are considered, and the functional and anatomical adaptiveness of the mature system is reviewed, as are new findings on the development of visual cortex. Of course, such an effort means that some topics have not been covered, or have not been covered adequately. In particular, we have omitted parts of the subcortical visual system that are not directly involved in influencing visual cortex and, thus, are not involved in visual perception. The visual hypothalamus, the pregeniculate nucleus, the visual sector of the reticular nucleus and the claustrum, the visual components of the basal ganglia, the accessory optic system, the pretectum, the superior colliculus, the visuomotor nucleus of the pons, and the visual cerebellum all deserve discussion, but this must occur elsewhere. Fortunately, a complete volume in this series is planned on the superior colliculus. Enjoy this introduction to the primate visual system.