Hallucinations are likely to occur if this system, either the visual system or the motor tracking system breaks down in some way.  In the absence of sufficient input, the brain will still be trying to complete its image and hallucinations may occur.

 

     When you look at how the cells in the eye and brain respond, you find specialization. Cells only respond to certain physical aspects of vision.  You can always tell this by impaling a cell with an electrode and carefully analyzing what the cell responds to.  For example in the retina when researchers examined ganglion cells with electrodes they found that the cells fired only under specific circumstances.  I have already alluded to the receptive field of a ganglion cell.  Some ganglion cells would only fire if an image appeared with a certain pattern, a light central area and a dark surround, or others would respond only in the opposite situation dark surrounded by light and so forth.  The rods and cones respond to light with a reduction of the so-called dark currents.  The rods and cones connect mostly with two types of cells.  There are the horizontal  cells that communicate with other rods and cones mostly.  Horizontal cells partly process and influence other receptor cells.  Many of them secrete an inhibitory chemical onto the membrane of rods and cones, GABA that in turn influences the currents.  This is complicated but horizontal cells are excited by primary visual receptors then end up doing almost the same thing that light impulse does, reducing the dark current, perhaps ending up by heightening or tuning in the response of a stimulated rod or cone light receptor.  Once a receptor is stimulated we are, more or less, further “tuned in” to it.

 

     Rods and cones connect most directly to the next line of cells the bipolar cells.  Bipolar cells also have no action potentials and instead have a graded or analog response.  The bipolar cells connect directly with the Ganglion cells.  The ganglion cells bring the message from the retina into the cranium and are the first digital or action potential, all or nothing type cells in visual processing.  However these in turn are influenced by the  amacrine cells.

 

[i]

Figure 1: The circuitry of the retina. Note that cells are more than relays. Signals are worked on and modulated before they are passed on. That is the function of the Amacrine and Horizontal cells. The more cells involved, the more a signal can be processed.

 

The ganglion cells extensions or axons gather into a cable that is the optic nerve.  The optic nerve carries visual impulses to two closely related but separate parts of the brainstem for processing.  In lower animals most of the fibers go to the midbrain the area of the superior colliculus, while in humans and higher forms, a separate area has developed from the same plate of nerve cells, the lateral geniculate nucleus of the thalamus*.  The thalamus is a way station to the ultimate termination of visual impulses in the cerebral occipital cortex. As we have seen in higher animals there is increasing cephalization of visual input.  For frogs, fish, and more primitive forms visual processing is largely retinal or at best in the dorsal midbrain but not in a cerebral cortex.  In man, these lower midbrain connections are preserved and serve some purpose, but the majority of visual input goes to that lateral geniculate nucleus in the thalamus and eventually to the cerebral cortex where visual stimuli are more extensively processed, i.e. reach conscious awareness. Let’s explore how this comes about.

 

EYE-BRAIN CONNECTIONS:

     While in humans and higher mammals, the optic nerve sends axons to the evolutionarily older midbrain tectum (roof), in lower animals this area is the ultimate terminus of visual processing. It is an area responsible for simple reflexive motor movements processed rapidly, mainly movements of the eyes themselves.  As we’ve seen, animals respond quickly to movements perceived visually, for the frog, to flick his tongue out at a fly, or the wildebeest to withdraw his neck from a lunge lion’s powerful jaw.  Movements perceived by rods in the peripheral retina have to be foveated, brought into more precise inspection by the central retina. An interesting thing happens when you are out at night looking at the stars. You get a glimpse of a faint object in the periphery of your vision. Then as you reflexly focus on it, drawing the star into your central gaze for inspection, it’s lost. Your more light sensitive rods in the retinal periphery perceive the dimly lit spot, but your cones in the central retina adapted for precise vision of well-lit objects, fail to respond to the dim image of a faint star.

 

     In the midbrain as elsewhere in the central nervous system, cells are organized into groups or nuclei.  These are anatomical aggregates such as the Superior Colliculus, meaning upper hill, which implies that there is a lower hill below or inferior colliculus which happens to be involved with auditory processing. The superior colliculus mediates some involuntary eye movements along with, a favorite for medical students in neuroanatomy, called the Edinger-Westphal nucleus.   Edinger-Westphal helps to control pupils as when the pupil constricts to light and you may notice the pupil also constricts when you accommodate to a near or close object.  In syphilis there is an Argyll-Robertson pupil that is small in size and does not constrict to light but will get smaller with accommodation.  It is like a prostitute it accommodates but doesn’t respond.  The midbrain tectum is involved with these and other reflexive but necessary eye responses and it is where the connections are made with the nerves that bring these reflexes about, most importantly the third or oculomotor nerve.

 

What we are describing here is the difference between simple reflex vision mediated by the brainstem in lower animal forms and its embellishment, perhaps the full appreciation of as visual scene, true perception rather than reflex motor function, enjoyed by more advanced brains.  The cortex contextualizes visual input and correlates vision with the other senses and experiences.

 

     In some animals the visual pathway essentially ends at reflex visual responses but no conscious embellishment. Humans expect more from our vision. In lower animals the midbrain tectum is the major terminus of visual input but in men, there is more crosstalk with other sensory modalities especially touch because touch fibers also project to this area.  In lower animals things are designed so their responses will be more automatic and rapid. In humans some of this projection of the optic nerve and midbrain areas is preserved performing various automatic adaptations of the eye, especially eye movements. A pinealoma an abnormal growth on the pineal gland can compress the superior colliculi of the midbrain (the pineal gland lies close to the midbrain collicui with cells of similar visual origin and light sensitive itself in lower animals).  This creates difficulty looking up with your eyes over the horizontal plane,  a “Parinaud’s syndrome”, the main sign of such a tumor.  A person with Parinaud’s syndrome would be unable to gaze upward enough to walk upstairs.

 

     The pineal gland that lies close to the mounds in the midbrain, the colliculi, referred to above, secretes melatonin, a hormone much written about in recent years that helps regulate sleep wake cycles.  Visual reflexes in part, also control melatonin secretion. It turns out that light impinging on the retina (as normally occurs during the day) blocks melatonin secretion and that darkness, conversely increases melatonin secretion. Melatonin may have many effects, but in part, it acts as a sedative, encouraging sleep.  Visual control of the sleep wake cycle again is likely far more important in lower animals than it is in man whose life is tied to artificial light situations and other regulators or inhibitors of sleep.

 

     The midbrain tectum connects to neurons from other brain areas that move the eye. Lower animals essentially have no visual cortex, but in man there are close connections in this area with cortical eye movement areas.  In the left and right frontal lobes are separate frontal eye fields that when stimulated push both eyes to the opposite side in a rapid eye movement (saccade).  In the occipital or visual back part of the brain can be found movement cells as well that move the eyes more slowly and help with visual following of a moving object.  When you are sitting in a moving train watching telephone poles fly past you fixate on one pole and as you pass the next pole your rods catch sight of it and jump to that one as the first falls out of view.  If someone is looking at you at the time he will see your visual fixation jump from one pole to the next.  There is a slow movement of following as you fixate on the first pole while the train continues to move, followed by a jumping saccade in opposite direction as you catch up with the second one. This is repeated as you catch one and the next pole in your passing train, a slow visual following movement followed by a saccade in the opposite direction repeated over and over again.  This is nystagmus and this specific kind is optokinetic nystagmus.  Both the frontal and occipital areas connect in man with the midbrain tectum in two groups of cortico-tectal fibers because they connect the cortex and the tectum. Eye movements in man are under increased cortical control as is visual reception so you can appreciate that there is cephalization of eye movement as well as reception.  In man and advanced animals everything is more cephalized.

 

 

.

 Figure 2[ii]: The pathway of visual images into the brain.  It is mostly because of the reversal of images in the retina that right field objects arrive into the left side of the brain and vice-versa. Upper field images go to the lower parts of the visual cortex.

 

 

 

     As animals evolved they developed higher connections that start with a more advanced diencephalon or thalamus and end by developing the cerebral cortex where we localize conscious feeling and perception and operations between percepts, association and then action. But it starts with the thalamus.  For vision, the vast majority of retinal ganglion fibers project to the lateral geniculate nucleus# .  The retinal ganglion cells project topographically or retinotopically.  The way they do so is rather complex.  Cells from the left side of the visual world go to the right lateral geniculate nucleus.  For the left eye retinal cells that pick up light from the person’s left are in the nasal half of the retina and project to the right lateral geniculate.  In other words nasal fibers cross. However, in the left eye, temporal fibers (recall that images are projected backward on the retina so temporal or outside fibers from the left eye receive vision from the right half of the visual field – see how objects project onto the retina in Figure 2), do not cross.  They go to the left lateral geniculate nucleus. Each lateral geniculate thus receives impulses from the opposite visual field,Y the right lateral geniculate gets impulses from the left side from both the right and left eyes.

 

THE BRAIN ORGANIZES VISUAL INPUT:

     Inside each lateral geniculate nucleus you can see a layering of cells, a lamination, under the microscope.  Fibers from each of the two eyes are kept separate and project to specific layers that are dedicated to one eye; nasal fibers from one eye and temporal fibers from the other are still separate within the nucleus. The cells then synapse with other cells inside the lateral geniculate and are still segregated retinotopically as they reach the primary visual cortex in the occipital lobe in the back of the brain.  The left hemisphere receives impulses from the right visual field, the right occipital lobe, receives fibers only from the right lateral geniculate sensitive to goings on in the left visual field. Certain brain neurons are hard-wired to receive specific inputs. This point was dramatically brought home in the classic experiments done by Roger Sperry.  He cut out the eyes of a frog, then reattached them through the optic nerves only upside down.  Most of the axons grew back to their original neuronal connections. What happened is that the frog, now seeing a fly on its upper left would jump or flick its tongue in the wrong direction, as to its mirror image!! Most disturbing for the poor frog but illustrating that neuronal connections respond to some specific stimulus. 

 

     Here we come to another universal in brain organization. We see a small spot of light in an area of our visual field.  That impinges on a tiny area of our retina that is connected to specific ganglion cells that project to specific (retinotopic) areas of the lateral geniculate nucleus.  This retinotopic organization is preserved as still a second set of cells in the lateral geniculate project onto the visual occipital cerebral cortex. You are lightly brushed on a tiny area on the right index finger. It turns out that little impulse is faithfully projected through a nerve, up the spinal cord to specific cells then up to the cerebral cortex where the touch receives some conscious awareness, All throughout this complex projection of the touch the topographic representation (this time somatotopic not retinotopic) is preserved.  If someone should stick you with a pin in the same place, an impulse would eventually arrive in almost the same location on the cerebral cortex. That is how you localize it and process it; that is how you know something happened in just this place on your finger. Impulses are faithfully projected in an organized manner to primary sensory cortices.  What will happen after that, where the sensation ends up may be very different.  A slight touch from someone you love is not the same getting jabbed with a pin at the same location. That will depend on further processing in secondary and tertiary cortical regions and also by some initial organization that separates sensory modalities.  The same holds true for the temporal areas involved with hearing.  These areas are tonotopic.  Neurons project in highly specific ways.   You can appreciate the homology between somatotopic organization for pain and touch over the body where the receptive field of a central neuron is a certain area over the body and tonotopic organization where a given neuron will only respond to a sound of a certain tone, and in vision, retinotopic organization where a neuron in that lateral geniculate or the occipital cortex will only fire when there is a visual stimulus on a certain region of the retina.  It is interesting that with smell (olfaction), one sensation that does not project through the thalamus there is no known topographic projection. But then there is nothing specifically spacial about smell apart from graduated increase or decrease of an odor’s intensity as you get farer or nearer to the odor-producing object.

 

     In the cortex then the right hemisphere processes visual input from you left visual field. That is another way of saying that the right hemisphere processes visual information from the right half of both the left and right retinas. Central or foveal vision is projected to both cerebral hemispheres.  There is extensive cross talking between hemispheres through the corpus callosum that interconnects them.  But as a general rule this retinotopic organization is preserved in the whole visual pathway from the retina to the occipital lobe. Parallel groups of axons are gathered into tracts of white matter, nothing more than big cables. From the lateral geniculate to the calcarine cortex where visual inputs are initially processed, we have the geniculocalcarine tract. This is a white matter bundle.

 

     We know that the geniculocalcarine tract carries retinotopic information and how the cortex receives it, mostly by observing the effect of lesions.  If a person has a large stroke in the back of the right hemisphere of the brain, he will not be able to see to his left side, a left hemianopsia. You can also monitor electrically the response of neurons in the right occipital cortex. In doing so you will find that first of all, the cortex is arranged in columns approximately 6 neurons deep.  The columns of cells continue to be retinotopic that is, it is composed of a group of cells that only respond to a visual stimulus from one area of the retina.  Most of the cortical cells no longer respond only to a single eye’s input, as do specific cells in the layers of the lateral geniculate. They are place coded. Not only that, many of them respond to a bar of light or a bar of darkness only in one specific orientation, that is only a certain number of degrees from the vertical.  These cells of the primary visual cortex that respond to stimuli of a certain orientation are simple cells.  Others respond only to various visual characteristics, for example, movement in a specific orientation or color and other features so that they are more specific with regard to movements and features and higher order of complexity therefore appropriately termed complex cells.

 

 

 

 

Figure 3:  PET scan three dimensional rendering of cortical activation of a visual image. Colored areas are activated areas of brain [iii].

 

     When you look at something there is a whole matrix of cells over the visual cortex, not only in the primary visual area that initially receives the input, but spread over a wide area of the cortex.  Many of these cells respond to specific superficial characteristics of the stimulus.  If you are looking at a nose of a person, the specific orientation of a vertical bar of light, color characteristics etc. And as you gaze over and explore the entire face, a huge group of specific cells is called into action, each responding to a fairly simple characteristic of the face, maybe as complex as the way the smile creases move as a person smiles.  We can appreciate how the whole scheme of seeing comes together, a merging of visual input and motor output to the eye.  We’ve talked about the limited field of view of precisely seeing cells, the cones in the fovea and the motor output to the eye allowing the eye to foveate over a whole visual scene.  This information is ultimately conveyed to the primary visual area in the occipital lobe then to secondary and tertiary visual areas of the brain.  The system is designed to respond to rather superficial properties of the visual stimulus.  Experiments have shown that these characteristics are even more widely distributed over the cortex. Not only do they ramify to visual areas of the brain, but also all sensory inputs are distributed multimodally, mixed with auditory, tactile and even olfactory images in widely separated regions of the brain.  By the time you have a picture of a friend, you have auditory associations with his voice and even emotional associations based on your arguments and good times etc. As the visual image interacts widely with an enormous storehouse of experiences and other sensory inputs the depth of the interaction is limited only by the complexity of cerebral circuitry and past experiences. No wonder that if our mind is set free to ramble we can come up with wide ranging ramblings and associations mirrored and perhaps caused by widely ramifying cortical representations of visual images.

 

HIGHER PROCESSING:

     I’ve discussed visual processing, starting from the initial effect on retinal rods and cones.  From the point of view of neurons in the cerebral cortex, there are basically two broad groups of connections.  The center of this activity we conveniently place at area 17 known as the primary visual cortex.  From this vantagepoint the two groups of connections are corticopetal (those going toward) and inter-cortical those fibers connecting the primary visual area to other cortical areas.  Corticopetal fibers we have already described.  These come primarily from the lateral geniculate bodies to the calcarine visual cortex and comprise the geniculocalcarine tract.  A lesion here, such as a stroke especially or a tumor will cause a restriction in visual fields because, as we have seen, fibers are segregated according to an area of the visual environment.  A total wipeout of fibers in the left side of the brain will cause loss of vision to a subject’s right side, a right hemianopsia. The lower fibers, going through the deep temporal lobe of the brain, carry visual impulses from the upper visual field so a lesion affecting just these fibers will cause a cut in vision in the right upper quadrant, a so-called “pie in the sky” defect called a quadrantanopsia.  Neurologists, ophthalmologists and other doctors spend a lot of time evaluating these visual field cuts because they are very obvious and localizing for lesions inside the brain but they are not that interesting or subtle.  The real challenge comes when you try to evaluate lesions affecting cortex-to-cortex connections.

 

     Cells that do simple visual processing aptly termed simple cells, and more Complex ones that are also hard-wired to respond to certain visual characteristics, Complex cells.  Then there are also hyper-complex cells.  The Retina of man projects to the back of the brain or occipital lobe (Figure) where all of this processing takes place.  There we find cells that are tuned (in other words respond by firing) to specific visual characteristics, for example there are cells that only respond to a light bar of a certain length surrounded by a dark area.  Our visual system is set up to increase contrasts and what the retinal ganglion cells respond to ordinarily are light areas  with dark surrounds and vice versa.  Again all of these tiny images are somehow integrated with specific borders built in to somehow form an image.  Actually we know little of how this image is finally put together and enters consciousness. 

 

     After an impulse arrives at the primary visual cortex, signals ramify to other cortical areas where they are further processed.  Defects in these intercortical pathways cause more subtle problems.  Many nerve fibers go from the calcarine cortex to the temporal lobe and are thus designated occipito-temporal. Part of the function of the temporal lobe is to provide cross talk between vision and other sensory modalities such as hearing and touch and it is here that visual imagery connects with language function. The language area of the brain is in the peri-Sylvian area of the left hemisphere in most individuals.

 

     Curious disorders occur with lesions in cortex adjacent to the primary visual area or most importantly, in white matter connections between these regions.  Visual scanning is integral to object recognition since the whole scene or image cannot be taken in a single instant. It can only be theorized that there is some kind of short-term memory evoked by scanning a scene to produce on the basis of small pieces of information, a total picture of the scanned object. This will further serve for recognition, say of a statue (as seen in the illustration) or a face for example.

 

     Destruction of intercortical fibers leading from primary visual cortex to the parietal lobes on both sides of the brain will cause Balint’s Syndrome. These are association fibers high in the back of the brain fairly close to the surface.  What you lose Is the ability to take in a whole scene, and have trouble keeping track of or recognizing more than one or two objects in a scene.  This is designated as simultanagnosia.  The condition is also characterized by inability to throw one’s eyes at a novel object in a scene, a sort of gaze stickiness.   We talked about seeing something new especially a moving novel object out of the corner of the eye in the domain of the rods, and our ability to bring the eyes to this new object (foveation). Seeing an object, we also have to be able to respond to it, to grab at it, touch it, even lunge at it. We see only a small part of a scene at any given moment.  Yet in specific situations,  a sword fight, or running on an open road,  we have the illusion that we have a picture of the whole scene.  Image storage and integration in the cerebral cortex accomplish that.  Thus that we can respond at any time with a reaching or motor movement to any part of the total scene. This is what is impaired in Balint’s syndromey Compare this with the figure of object scanning (Error! Reference source not found.). This may or may not be associated with a cut in one’s visual field, depending on the extent or depth of a lesion, but is most often associated with an inferior quadrant field cut. The interesting thing is that a person may well not be able to recognize an object on account of an inability to form a whole picture of it not because of any language problem or trouble actually seeing it. He may well be able to recognize something in only a small part of his visual field but not an object that takes up a whole field.

[iv]

Figure 4: In this patient  metastases destroying  both parieto-occipital junctions high in the brain caused a "Balint's Syndrome".

 

     Problems in connections between the primary calcarine receiving visual cortex and other areas of the brain that do more advanced and multimodal visual processing have been great neurological curiosities for a long while.  Some patients see easy enough, but can’t find their way around and unless they literally memorize their environment consciously feature by feature.  They must know for example, that there are exactly six steps before coming to a corner of a wall or a chair before they are to make that right turn to find the way to their room or they will get lost.  You and I don’t find our way around like that.  We depend on a referential inner map of our environment that is subliminal, barely noticed and automatically constructed, in other words a facility for recognition of the familiar. A person, unable to develop a total picture somehow, suffers from a sort of environmental agnosia or toporgraphagnosia.,  This may be called by different names but lesion-wise have something very similar to a Balint syndrome, they are unable to orient within a total picture of the environment and depend on the individual tree in a forest but have only a very poor concept of a forest.

 

     Color recognition is simpler but is much written about.  Some people can’t recognize or name colors or of find named colors and so have a problem with color recognition.  This problem happens with disconnection between the inferior occipito-temporal fibers that connect vision with the language area of the brain. You can lose your color reception just in one visual field, to the right typically, or on both sides.  A right sided visual field cut is associated a lot of the time because the right visual field registers in the left hemisphere.  Language function resides in the left brain in most of us, and so it not unusual for the connecting fibers that go from the language area to the left temporal lobe to be affected by a lesion in the back of the left hemisphere. 

In prosopagnosia a person has a problem recognizing faces.  One is unable to  distinguish faces of friends and relatives (though he may be able to tell them apart by listening to their voice) or even to distinguish objects within a category. He may have to find his car in a parking lot by searching all license plates rather than by just recognizing its appearance. A farmer may not be able to tell on of his farm animals from another.  With prosopagnosia there are usually lesions in both the left and right occipito-temporal intercortical pathways.

 

     What are we seeing here?  Disorders of visual recognition involve a higher order of abstraction than simple unprocessed vision. The visual image arrives at the primary occipital receiving station is processed here but then higher order processing and relationships are quickly handled by other areas of the brain, initially adjacent but ultimately far-removed from the calcarine cortex.  We know this in many ways.  You can follow a change in electrical activity evoked by simple and by complex visual stimuli over the surface of the brain. This is difficult because ordinarily such electrical potentials are small, only a small fraction of a microvolt in size and the electrodes used are far from the brain where such potentials arise, if they are ordinary electrodes placed over the surface of the skull.  But such evoked responses can be observed when the skull is open during neurosurgery at which time you can place electrodes directly over the surface of the brain.  This is rarely done except in epilepsy surgery where arrays of electrodes are placed in order to find the electrically active discharging focus of activity.  Also during surgery, you can put a probe over the brain surface and see what sort of experience the electrical probe evokes stimulating a specific anatomical spot over the brain.  This was originally done rather crudely many years ago by Wilder Penfield a Canadian neurosurgeon.  The patient needs to be awake and this is rarely done anymore.  T

To Part III

www.pneuro.com

 

 

     The PET (Positive Emission Tomography) scanner watches the utilization of glucose, in other words the metabolism or activity of brain regions. You can watch various brain regions metabolizing glucose as they become active using color-codes and use of various individual parts of the brain and charts the activity or glucose uptake over the brain’s surface. As a person looks at or scans an object you can see what parts of the brain become activated and the sequence of anatomical activation.  With that technique you see that even a simple visual object or scene, quickly activates a wide area of brain.  Over the years the most important information has been culled by examining patients with certain known cerebral lesions to give a picture of localization of function and the understanding of connections within the brain.  Neuroscientists have gradually evolved a picture of brain function that is modular.  Certain brain regions perform a specific given function.  There is primary visual, primary auditory, primary somateshtetic, olfactory, etc cortex and a language area of the brain.  Each of these areas has to perform a certain function and also be connected to the others.  Each are gray matter areas containing neurons.  They communicate using white matter tracts or bundles of axons. When a problem is detected there may be an anatomical defect in neurons or connecting axons.  For example, a person may be able to see and name letters, he may perform just perfectly on a Snellen eye chart with good visual acuity,