{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Number 8 HUBEL - The Visual Cortex of the Brain by David H...

Info icon This preview shows pages 1–9. Sign up to view the full content.

View Full Document Right Arrow Icon
Image of page 1

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
Image of page 3

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 4
Image of page 5

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 6
Image of page 7

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 8
Image of page 9
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: The Visual Cortex of the Brain by David H. Hubel November 1963 A start toward understanding how it analyzes images on the retina can he made through studies of the responses that individual cells in the visual system of the cat give to varying patterns of light 11 image of the outside world striking A the retina of the eye activates a most intricate process that results in vision: the transformation of the reti- nal image into a perception. The trans formation occurs partly in the retina but mostly 'in the brain, and it is, as one can recognize instantly by considering how modest in comparison is the achievement of a camera, a task of impressive mag- nitude. The process begins with the responses of some 130 million light-sensitive re- ceptor cells in each retina. From these cells messages are transmitted to other retinal cells and then sent on to the brain, where they must be analyzed and inter- preted. To get an idea of the magnitude of the task, think what is involved in watching a moving animal, such as a horse. At a glance one takes in its size, form, color and rate of movement. From tiny differences in the two retinal images there results a three-dimensional pic- ture: Somehow the brain manages to compare this picture with previous im- pressions; recognition occurs and then any appropriate action can be taken. The organization of the visual system —a large, intricately connected popula- tion of nerve cells in the retina and brain —is still poorly understood. In recent years, however, various studies have be- gun to reveal something of the arrange- ment and function of these cells. A decade ago Stephen W. Kufiler, working with cats at the Johns Hopkins Hospital, discovered that some analysis of visual pattems takes place outside the brain, in the nerve cells of the retina. My colleague Torsten N. W'iesel and I at the Harvard Medical School, exploring the first stages of the processing that occurs in the brain of the cat, have mapped the visual path- way a little further: to what appears to be the sixth step from'the retina to the cortex of the cerebrum. This kind of work falls far short of providing a full understanding of vision, but it does con- vey some idea of the mechanisms and circuitry of the visual system. In broad outline the visual pathway is clearly defined [see bottom illustration on following page}. From the retina of each eye visual messages travel along the optic nerve, which consists of abOut a millipn nerve fibers. At the junction known as the chiasm about half of the nerves cross over into opposite hemi- spheres of the brain, the other nerves re- maining on the same side. The optic nerve fibers lead to the first way stations in the brain: a pair of cell clusters called the lateral geniculate bodies. From here new fibers course back through the brain to the visual area of the cerebral cortex. It is convenient, although admittedly a gross oversimplification, to think of the pathway from retina to cortex as con- sisting of six types of nerve cells, of which three are in the retina, one is in the geniculate body and two are in the Certex. Nerve cells, or neurons, transmit mes- - sages in the form of brief electrochemi- cal impulses. These travel along the outer membrane of the cell, notably along the membrane of its long principal fiber, the axon. It is possible to obtain an electrical record of impulses of a sin- gle nerve cell by placing a fine electrode near the cell body or one of its fibers. Such measurements have shown that im- pulses travel along the nerves at veloc- ities of between half a meter and 100 meters per second. The impulses in a given fiber all have about the same am- plitude; the strength of the stimuli that give rise to them is reflected not in ampli- tude but in frequency. At its terminus the fiber of a nerve cell makes centact with another nerve cell (or with a muscle cell or gland cell), forming the junction called the synapse. At most synapses an impulse on reaching the end of a fiber causes the release of a small amount of a specific substance, which dilluses outward to the membrane of the next cell. There the substance either excites the cell or in- hibits it. In excitation the substance acts to bring the cell into a state in which it is more liker to “fire"; in inhibition the substance acts to prevent firing. For most synapses the substances that act as transmitters are unknown. Moreover, there is no sure way to determine from microSCopic appearances alone whether a synapse is excitatory or inhibitory. It is at the synapses that the modi- fication and analysis of nerve messages take place. The kind of analysis depends partly on the nature of the synapse: on how many nerve fibers converge on a single cell and on how the excitatory and inhibitory endings distribute themselves. In most parts of the nervous system the anatomy is too intricate to reveal much about function. One way to circumvent this difficulty is to record impulses with microelectrodes in anesthetized ani- mals, first from the fibers coming into a structure of neurons and-then from the neurons themselves, or from the fibers they send onward. Comparison of the behavior of incoming and outgoing fibers provides a basis for learning what the structure does. Through such exploration of the different parts of the brain con- cerned with vision one can hope to build up some idea of how the entire visual system works. hat is what W'iesel and I have under— taken, mainly through studies of the visual system of the cat. In our experi- ments the anesthetized animal faces a wide screen 1.5 meters away, and we shine various patterns of white light on the screen with a projector. Simultane- RETiNAL GANGLION STRUCTURE OF RETINA is depicted schematically. Images fall on the receptor cells, of which there are about 130 million in each nation. Some analysis of an image occurs as the receptors trans- mit messages to the retinal ganglion cells via the bipolar cells. A RETINA --——‘|——- _'.-..,-‘ -....___.A........-_.___-___.._.._.. «I .......___—___—_..i..._.—-..._—————-uu-‘I __._.._--.-—-. -f BIPOLAR RECEPTOR CELLS _ ‘ .' -. OPTICNEPVE: J' group of receptors funnels into a particular ganglion cell, as indicated by the shading; that group forms the ganglion cell‘s re- ceptive field. Inasmuch as the fields of several ganglion cells over- lap. one receptor may send messages to several ganglion cells. COMPLEX CORT—CL; CEL: .——fi' '——'-'-5-'-_____ __-_.____...—— ——’ .. ._ ____ _.. 9/ — VISUAL PROCESSING BY BRAIN begins in the lateral genicu- late body, which continues the analysis made by retinal cells. In the cortex “simple” cells respond strongly to line stimuli, pro- vided that the position and orientation of the line are suitable for a particular cell. “Complex” cells respond well to line stimuli, but 0 the position of the line is not critical and the cell continues to rospond even if a properly oriented stimulus is moved, as long as it remains in the cell‘s receptive field. Broken lines indicate how receptive fields of all these cells overlap on the retina; solid "I129: how several cells at one stage afiert a single cell at the next stage. ously we penetrate the visual portion of the cortex with microelectrodes. In that way we can record the responses of in- dividual cells to the light patterns. Sometimes it takes many hours to find the region of the retina with which a particular visual cell is linked and to work out the optimum stimuli for that cell. The reader should bear in mind the relation between each visual cell—no matter how far along the visual pathway it may be—and the retina. It requires an image on the retina to evoke a mean- ingful response in any visual cell, how- ever indirect and complex the linkage maybe. ' The retina is a complicated structure. in both its anatomy and its physiology, and the description I shall give is high- ly simplified. Light coming through the lens of the eye falls on the mosaic of re ceptor cells in the retina. The receptor cells do not send impulses directly through the optic nerve but instead con~ nect with a set of retinal cells called bipolar cells. These in turn connect with retinal ganglion cells, and it is the latter set of cells, the third in the visual path- way, that sends its fibers—the optic nerve fibers—to the brain. This series of cells and synapses is no simple buclret brigade for impulses: a receptor may send nerve endings to more than one bipolar cell, and several receptors may converge on one bipolar cell. The same holds for the synapses between the bipolar cells and the retinal ganglion cells. Stimulating a single re ceptor by light might therefore be ex- pected to have an influence on many bipolar or ganglion cells; conversely, it should be possible to influence one bipolar or retinal ganglion cell from a number of receptors and hence from a substantial area of the retina. The area of receptor mosaic in the retina feeding into a single visual cell is called the receptive field of the cell. This term is applied to any cell in the visual system to refer to the area of retina with which the cell is con- nected—the retinal area that on stim- ulation produces a response from the cell. Any of the synapses with a particular cell may be excitatory or inhibitory, so that stimulation of a particular point 011 the retina may either increase or decrease the cell's firing rate. More- over, a single cell may receive several excitatory and inhibitory impulses at once, with the result that it will respond according to the net eEect of these in- puts. In considering the behavior of a single cell an observer should remem- ber that it is just one of a huge popu- 21 _ ‘ _ I ‘1',”qu "1-,”. - :T‘f- _. lawm‘f‘fl'tfi '._flfi','.‘il‘llll<'liifr’-".‘|'l-?TI rI - I . ”$341.35? ii." h" .. _. ;.: ' . l .a , ' a n In)” 7/. 4". . CEREBELLUM POSTLATERAL GYRUS ‘ LATERAL GYRUS CORTEX 0F CAT’S BRAIN is depicted as it would be seen from the top. The colored region indicates the cortical area that deals at least in a preliminary way with vision. CHIASM LATERAL GENICULATE BODY OPTIC NERVE VISUAL SYSTEM appears in this representation of the human brain as viewed from be- low. Visual pathway from retinas to cortex via the laleral geniculale body is she“ n in color. 22 I ] BIOLOGICAL AND DEVELOPMENTAL DETERMINERS OF BEHAVIOR ' LIGHT lation of cells: a stimulus that excites "ON"'CENTER F‘ELD ' -- one cell will undoubtedly excite many others, meanwhile inhibiting yet another array of cells and leaving others entirely unafiected. For many years it has been known that retinal ganglion cells fire at a fairly steady rate even in the absence of any stimulation. Kulller was the first to ob- serve how the retinal ganglion cells of mammals are influenwd by small spots of light. He found that the resting dis- charges of a cell were intensified or di- minished by light in a small and more or less circular region of the retina. That region was of course the cell’s receptive field. Depending on where in the field a spot of light fell1 either of two response could be produced. One was an “on” response, in which the cell's firing rate increased under the stimulus of light. The other was an “OE” response, in which the stimulus of light decreased the cell’s firing rate. Moreover, turning the light 0E usually evoked a burst of impulses from the cell. Kufiler called the _ retinal regions from which these re- sprmses could he evoked “on” regions _M and“ofi'regions. 0n mapping the receptive fields of a CONCENTRIC FIELDS are characteristic of retinal ganglion cells and of geniculate cells. large number of retinal ganglion cells At top an oscilloscope recording shows strong firing bv an “on"{enter tvpe at cell when a - u w u . - - . . . to (1 El” re as, Kuiil dis- spot of light strikes the field center; ii the spot hits an “Oil" area, the firing is suppressed 381:: $3": tlclere 3;; two £2tinct until the light goes 03. At bottom are responses oi another cell of the "ofl"-oenter type. cell I In one the I live field consisted of a small circular "on" area and a surrounding zone that gave "oil" responses. Kufier termed this an “on”- oenter cell. The second type, which he called “oli"-center, had lust the reverse form of field—an “OE” center and an “on” periphery [see top illustration on this page]. For a given cell the effects of light varied markedly according to _ the place in which the light struck _’ . the receptive field. Two spots of light shone on separate parts of an “on” area ‘ produced a more vigorous “on” response than either spot alone, whereas if one spot was shone on an "on” area and the other on an "oil" area, the two efiects tended to neutralize each other, result- ing in a very weak “on” or “off” response. In an “of-center cell, illuminating the entire central “on” region evoked a maximum response; a smaller or larger spot of light was less effective. Lighting up the whole retina diflusely, even though it may aEect every receptor in the retina, does not direct a retinal ganglion cell nearly so strongly as a small circular spot of exactly the right size placed so as to cover precisely the SIMPLE cannon. onus have receptive and; or varietus types. In all or them the recepfivefield center- The min 0011- “un” and “all” areas, represented by colored and gray dots respectively, are separated Gem Of these 00115 seems to be the 0011' by straight boundaries. Orientations of fields vary, as indicated particularly at a and b. Wt in illumination between 0119 11351131 In the cat’s visual system such fields are generally one millimeter or less in diameter. region and surrounding regions. HUBEL | THE VISUAL CDRTEX OF THE BRAIN 23 Retinal ganglion cells differ greatly in the size of their receptive-field cen- ters. Cells near the fovea (the part of the retina serving the center of gaze} are specialized for precise dis- crimination; in the monkey the field centers of these cells may be about the same size as a single cone—an area sub- tending a few minutes of are at the cornea. On the other hand, some cells far out in the retinal periphery have field centers up to a millimeter or so in diameter. (In man one millimeter of retina corresponds to an arc of about three degrees in the ISO-degree visual field.) Cells with such large receptive- ficld centers are probably specialized for work in very dim light, since they assronsa ls WEAK when a circular epel at light a shone on receptive field or . simple can sum UP messages from a large num- cortical cell. Such spots get a vigorous response from retina] and genicnlate cells. This ber of receptors. cell has a receptive field of type shown at a in bottom illustration on opposite page. Given this knowledge of the kind of visual information brought to the brain by the optic nerve, our first prob- lem was to learn how the messages were . handled at the first central way station, the lateral geniculate body. Compared with the retina, the geniculate body is a relatively simple structure. In a sense there is only one synapse involved, since the incoming optic nerve fibers end in cells that send their fibers directly to the visual cortex. Yet in the cat many optic nen‘e fibers converge on, each geniculate cell, and it is reasonable to expect some Change in the visual mes- sages from the optic nerve to the genicu- late cells. When we came to study the geniculate body, we found that the cells have many of the characteristics Kuiller described for retinal ganglion cells. Each genicu- late cell is driven from a circumscribed retinal region (the receptive field) and has either an "on" center or an “OE” center, with an opposing periphery. There are, however, differences between geniculate cells and retinal ganglion cells, the most important of which is the greatly enhanced capacity of the periph- ery of a geniculate cell's receptive field to cancel the eEects of the center. This means that the lateral geniculate cells must be even more specialized than reti- nal ganglion cells in responding to spa- tial differences in retinal illumination rather than to the illumination itself. The lateral geniculate body, in short, has the function of increasing the disparity—al- ready present in retinal ganglion cells— between responses to a small, centered spot and to difluse light. In contrast to the comparatively sim- Ple lateral geniculate body. the cerebral IMPORTANCE or ORIENTATION to simply oortical'eells is indicated by varying cortex is a structure 0‘ stupendous com- responses to a slit of light from a cell preferring a vertical orientation. Horizontal slit plexity. The cells of this great plate of {mp} Produces no response, slight tilt in weak response, vertical slit a strong response. ‘ 24 COMPLEX CORTICAL CELL responded vigorously to slow down- ward movement of a dark, horizontal bar. Upward movement of bar produced a weak response and horizontal movement of a gray matter—a structure that would be about 20 square feet in area and a tenth of an inch thick if flattened out— are arranged in a number of more or less distinct layers. The millions of fibers that come in from the lateral geniculate body connect with cortical cells in the layer that is fourth from the top. From here the information is sooner or later disseminated to all layers of the cortex by rich interconnections between them. Many of the cells, particularly those of the third and fifth layers, send their fibers out of the cortex, projecting to centers deep in the brain or passing over to nearby cortical areas for further processing of the visual messages. Our problem was to learn how the infor- mation the visual cortex sends out dif- fers from what it takes in. Most connections between cortical cells are in a direction perpendicular to the surface side-to-side connections are generally quite short. One might there- fore predict that impulses arriving at a particular area of the cortex would exert their effects quite loeally. Moreover, the retinas project to the visual cortex (via the lateral geniculate body) in a systematic topologic manner; that is, a given area of cortex gets its input ulti- mately from a circumscribed area of retina. These two observations suggest that a given cortical cell should have a small receptive field; it should be in- fluenced from a circumscribed retinal region only, just as a geniculate or retinal ganglion cell is. Beyond this the anatomy provides no hint of what the cortex does P“—_‘" --—.._.__ —————-.‘ L__——_-.—__————__.-..___J with the information it receives about an image on the retina. In the face of the anatomical com- plexity of the cortex, it would have been surprising if the cells had proved to have the concentric receptive fields characteristic of cells in the retina and the lateral geniculate body. Indeed, in the cat we have observed no certical cells with concentric receptive fields; instead there are many diiierent cell types, with fields markedly dilterent from anything seen in the retinal and geniculate cells. The many varieties of cortical cells may, however, be classified by func- tion into two large groups. One we have called “simple”; the function of these cells is to respond to line stimuli—such shapes as slits, which we define as light lines on a dark background; dark bars (dark lines on a light background), and edges (straight-line boundaries between light and dark regions) . \Vhether or not a given cell responds depends on the orientation of the shape and its position on the cell’s receptive field. A bar shone vertically on the screen may activate a given 'cell, whereas the same cell will fail to respond [but others will respond) if the bar is displaced to one side or moved appreciably out of the vertical. The second group of cortical cells we have called “complex”; they too respond best to bars, slits or edges, provided that, as with simple cells, the shape is suitably oriented for the particular cell under observation. Complex cells, how- vertical bar produc...
View Full Document

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern