About science; About vision
This is about science produced by the California Institute of Technology and originally broadcast by station KPCC Pasadena California. The programs are made available to the station by national educational radio. This program is about vision with host Dr. Peter less a man and his guest Dr. Derrick Fandor. Here now is Dr. listen. Have you ever taken the time to think about that old phrase the mind's eye. Of course what it means is the ability to remember something in picture form to see something as ones who saw it. But that phrase comes close to telling us how things really are. The eye without the brain's ability to translate what it sees is useless while the mind without the eyes ability to recognize and tell about light is limited. So we see that this business of seeing is a cooperative venture that works so well for most of us that we seldom realize what's going on. Scientists have known for years of course that the eye is a device for
detecting light changing the light it receives into coded nerve impulses and sending them along the optic nerve to the brain. But what are these impulses. Are they tiny pictures long messages or maybe bits and pieces of numbers similar to what would work in a computer. That is what some of the latest work is being aimed about specifically at how the brain gets what it gets from the eye what form the information is in where it goes in the brain and what the brain does with the information across CVS. Here to discuss these problems with us now is Dr. Derek Fender professor of biology and Applied Science accounting. Dr. fender and his colleagues have been working with a very small electric signals generated by the optic centers of the brain when the eyes see flashing lights. Dr. fender a native of Heath England obtained his bachelor's degree at Reading University in 1939 and his Ph.D. in physics at the same university in 1956.
He was a senior lecturer at the university before joining the Kel-Tec faculty in 1961. Derek friend has done research on the function of eye movements in the visual process and is now engaged in biological systems analysis. Derek What exactly does and how does the brain analyze visual information and recognize what the eyes see. Well before before I get into that problem itself I would like to give due credit to a colleague of mine who works with me in this field. This is Dietrich Lehmann who is a neuro psychiatrist who actually works at the Institute for visual Sciences in San Francisco. We run this as a cooperative research project. So you are really speaking just as much for himself or for here as for yours you must blame him for at least half of what I thought you were being very noble in sharing the credit for to have already taken the way out.
As for the problem itself this is really a very big problem. Just how the human brain analyzes the visual information and enables us to recognize specific things in the visual field. There are many people in fact working many research teams working on this all over America and in fact all over the rest of the world and they're attacking this problem at various levels and you'll see what I mean by levels in a few moments I hope we ourselves are interested in one very small level of this information chain and that is what actually happens. Do we believe visual information as it makes its way from the eye to the visual cortex of the brain. Now that in fact is a long path. If if I can just trace this the
passage of the information over this path with you it goes something like this. Jerry could I interrupt for a second to ask as a very simple person what is the cortex to which this message is transmitted and the cortex is specifically the surface layers of the brain or the surface layers of nerves. And we usually refer to reserve the term cortex for these layers of nerve cells which respond to primary sensory input. That is if you like goes into some visual cortex like an active sound in the ear then some cells in the auditory cortex become active and so on. But these are all surface layer cells not the deep seeded cells which perform more fundamental processes of the thinking and knowing and making abstractions proving theorems. And so on.
So this is this is sort of the entry point it's like the mailbox of the brain that's where the messages come in first the cortex is really the first point at which the brain receives information from the outside world. Yes this is quite true. Now if I can get back to my to my long or long route the like from the outside world goes in through the eye through the pupil and through the lens of the eye in his image on the retina or the simple textbooks have pictures of cameras at this point and compare the eye with a camera. And in fact it behaves much the same as a camera up to that point. That is the image is upside down and backwards and so on in the psyche. The retina images formed on the retina of the retina is at the back of the eye inside the eyeball. And it is a mosaic of retinal receptors. There are the hundred million or so of these in each
eye. They have visual pigments in them. The pigment absorbs the light and in the act of absorbing the light it changes the light signal into an electrical signal. A very small electrical signal which flows off as a nerve impulse and this is the first point at which a change is being made. But from there on the light and the picture the visual scene has been lost. The brain only has nerve impulses little pulses of electricity on which to do its work. The amount of information encoded by the retina such as a small only twelve brightness and color probably the code brightness and color into nerve impulses and usually the coding is made by the rapidity at which these pulses follow one after the other. The nerve impulses then flow off from the retina to the brain
through a big bundle of nerves about a million there was in the bundle called the optic nerve. This nerve makes its way from the two are used with nerves make their way in place from the two ears through the quiet ASM which is just behind the eyes underneath the brain which some of the nerves cross over but some of the nerves do not on their way to the brain. You remember of course of the brain is split into two hemispheres left and right hemisphere. The crossing of the nerves in the Kiowas is such that if you were to look exactly straighter and or move the visual information which is out to your left hand side finishes in the right hand hemisphere of the brain and all that which is up to the right hand side finishes in the left hemisphere of the brain. This is of course quite typical of a human but the control the sensory input is
crossed and the control from the brain to the limbs is crossed. Left side of the brain controls the right hand and so on. Do we have any idea why that is so. Why you have a crossing is necessary. You know that's a very difficult problem to answer. If you follow up through the development of the various animals you find that that the insect level there is no crossing left his left and right is right. But they don't get mixed up as you go up to higher levels. Certainly by the time you get to things like frogs then there is a crossing of some information but other information is not crossed. And finally by the time you get to things as advanced as cats say then this crossing has occurred. But at what stage in the life or in the evolution of the animal kingdom. The decision was made to cross the hemispheres completely like this is really very difficult to say and the reason for which
the crossing occurred so anyhow the information passed through the chi or somewhere it's crossed and. Essentially left goes in to right and right goes to the left. And this information finishes in the cortex. We've been talking about before but specifically in the visual cortex and the cortex itself is quite large. It is as many subdivisions. They all have silly Latin names but also fortunately have numbers and we are talking specifically today about areas 17 18 and 19 of the visual cortex. I was 17 18 and 19 are a lot of the back of your head if you know so interlaced your fingers and put your hands behind your head in a comfortable relaxed gesture the palms of your hands would be neatly fitting over the area 17 in each hemisphere of your heart of your brain.
OK well now that is the first relay station in the brain as it were. These are the primary projection in the areas where sensory stimuli first become available to the brain itself and the point at which the brain can first start working on the little electrical impulses which are representing the visual scene from the outside world fact that the brain still has to make something of this message. Yeah that is true the perception has to go on perception that is converting it into a picture which we see in the mind's eye you that you talked about earlier on perception goes on in other areas of the brain which I could just call the higher centers the higher levels of the brain. And my work specifically today stops at areas 17 18 19. And perception in higher levels is another problem not only another problem for me but another problem for other people who are not wonder what about the coding
mechanism that that is used. Well. This of course brings us really to the next big problem. Just what sort of information is transmitted from the outside world to the cortex. I have said that the retinal intercept is really only sample brightness and color in the visual world and it would be possible just to transmit this information and no more to the cortex and then to use the rest of the brain as a big computer like make an ism to figure it all out and figure it all out. Yes there are other ways of doing it however and the one of the other way is to start. Making abstractions of the visual field as far forward in the visual system as you can I mean as close to the to the eyeball as you
can by abstractions I mean things like straight lines edges movement. It would be perfectly possible to detect the use parameters of the scene. And rather than transmit all of this hash about brightness to transmit the information which say it is there is a straight line in this part of the visual field. This is a bank tree produces a reduction of information and many conserved. Capacity in the nerve fibers and in the brain when the brain has to perform its final analysis of the picture. So in the characteristically scientific way you strive to use the simplest of model in the simplest of our impasse to find out how the system works. Yes that is true although it's very difficult to
say what is simplest in this case. You see this coding principle has been examined on animals and animals you actually open up the brain of the animal you stick an electrode into a cell in area 17 in the cortex and then you run around in front of the animal and wave a straight line in his visual field until eventually if you have chosen the right cell in the cortex you find that that cell fires when the straight line is in the right position. No straight lines seem to be pretty simple but are really pretty simple to one animal. In fact the work which has been done on animals which has ranged from frogs rabbits pigeons kept monkeys and so on is a whole slew of these experiments. The work which is being done on animal shows that some of the Hiram walls monkeys tend to analyze things like straight lines right down at the other end of the scale. Frogs
tend to analyze things which are of interest to them in a very real sense for example the frog is thought to have a mechanism which detects meal worms wriggling in its visual field. Now it relies on meal worms for its food and it has specific cells in its brain. There are just to detect meal worms. Now here you have problems deciding which is simplest if you were a geometer straight lines are simple. If you're a hungry frog meal worms are simply sleep and this is one of the big problems in this sort of work to decide for what sorts of stimuli you should look whether you are going to look for things which are geometrically simple or behavior at least simple. I mean have great meaning in the life of the animal. And how does this relate to the human visual system. Why don't we would like to ask exactly the same
question about the human does the human have straight lying detectors. Does he have middle worm detectors and so on. Or does the human perform his visual analysis in quite a different fashion. In fact we would like to ask the question does the human have exactly the same coding mechanisms. Unfortunately you can't do the experiments in the same way that if you were trying to take the back of the head of the human and stick electrodes in his cortex he gets upset about this to say the least. That's kind of disturbing his civil rights and I would accept not only your civil rights. So we have to adopt other tricks other experimental techniques but essentially we are trying to answer the same question as has been solved already for a large number of animals. But as you say with a human being you can't really take it to pieces and open it up and look in. So
what are some of the sort of things you can do there. Well of course your experiments on humans essentially have to be made from UPS like the wind. You can make experiments which just involve asking the human questions. Although when you rely on the truth of the human. And in answering them and we prefer other routes if possible. Not that we have any doubts about the truthfulness of our subjects but we like to avoid the possibility that the root which we use to record the brainwaves of these subjects the electroencephalogram or the we as it's called. And we argue rather like this. Suppose that the human does have. Nerve cells in his visual cortex which are there specifically to analyze straight lines. Then when straight lines of shame to the subject these nerve cells will become active. They when they're
active they produce little pulses of electricity. If we put electrodes on the surface of the hit over the area 17 then these electrodes should be able to see some small amount of that electrical activity. In fact the electrical activity is very small and the signal which you see on the head is or million for the vault or something like that very very small indeed. And what is more this activity of our hypothetical selves which are detecting straight lines is buried hopelessly beneath the electrical activity of all the other processes going on in the heat all the thinking and spontaneous activity which accompanies us all sort of wondering how long the silly doctor is going to go on fooling around with you. That's right. Yes yes. Well we get over that by adopting some of the
standard techniques which have been worked out in many other disciplines for detecting signal in noise. Where you show these straight lines not as continuous straight lines but as a flashing signal so that if there are populations of neurons which detect straight lines we will flash them into synchronous activity with the light and then we can average out the ongoing activity by chopping up the signal which we get into lengths which are exactly equal to the flashing rate of the light. Averaging all of these signals together. The synchronous activity will stay behind in the averaging process and the ongoing activity averages out. This is one of the standard. Tricks for detecting signal noise in practice we have to be a little more sophisticated than that because of the ongoing activity
is not truly noisy. It does it does add in in some ways and we have to use various filtering techniques but it can be done. That's important point and this signal which we're looking for which is only a few million of about varied between us but meter signal which is maybe 10 times bigger can be saved can be dug out of this hopeless welter of ongoing brain activity. I think we should make clear that Derek when you talk about noise you don't mean the thing that most ordinary people think about as something that affects the snow I'm talking in tiny engineering terms of course of the moment when things which you don't want in any measurement is court not used. And that's just the jumble of mixed up signals of all the different things that are going on. Let me write you talk always about straight lines. Is this a very. Very prototype of our visual object that you give people
or going to the other thing. In fact we do use straight lines as our stimuli very often we use targets with straight lines in them and targets with patterns of random dots patterns of curved lines patterns having a well marked series of contours patterns with various colors patterns having binocular disparity different patterns for the two RUC and so on because we are at the moment in the exploratory stages of this work really. It's been going on for many years but it still has many years to stretch into the future and we are still trying to find those patterns to which the human is most sensitive. So really all that we can do at the moment is to make a guess. The patterns which the human might like to see if he analyzes as a
geometer Or maybe those patterns which the human might like to analyze. If he behaves or he analyzes in a behavioral sense when the analog of the frogs. Yes you mean what is a wriggling meal worm to a man. Yes I mean does the human have a visual detector for a rare steak. It's a it's a far cry and I don't think it does that. But one ought to explore all of that field as well. You see in what we just haven't got around to all of these possibilities yet. Yes I suppose the real point is that that human most frogs like mealworms but some humans like steaks and other like lovers like pretty girls and I don't like racing sloops. Yes well if you take this argument too far and I suspect you might even have us off the air. OK well the upshot of this is then that we get this brain wave this is evoked potential as we call it because it is
evoked by a flashing light. And when we have extracted it from the noise we find goods move where you form it has well marked peaks and well-marked drops and it dips and bumps with some major peaks which are always there from one subject to the next highly reproducible with a lot of little wiggles superimposed on these big major waves. And the important thing is that some of these peaks we found are sensitive to the various stimuli which we have tried to that is we find in one of these evoked potentials a peak which if we show more and more and more straight lines the peak in your blood potential becomes bigger and bigger and bigger. In other words you have identified the particular thing that responds to to straight line. Yes.
So we we identified this and. This particular peak with the activity of a population of neurons whose job is specifically to analyze straight lines we can find peaks which analyze binocular disparity. These were the signals which we need for stereoscopic for depth perception. We can find peaks which are responsive to color and we can find peaks which are responsive to patterns of dots having different statistics. So with that they generate their texture. You see close dots round or widely spaced dots changing the texture of the picture so we can find peaks which are sensitive to textures. But we haven't gone on to the rare steak because of the stimulus at the moment.
The problem which we have of them under the analysis is that although we can find these peaks and therefore these populations of neurons in almost all of our subjects the peaks don't always occur at the same place along the way form. And this we interpret as meaning that we detailed the structure of the cortex the actual way in which each individual wires up his cortex is broadly the same but is different in detail. Maybe this is a developmental process that when we are born we just have the cortex wired up to detect a few primitive things like straight lines and so on. And then in the first few months of our life extra connections occur in the cortex wiring the cells up to detect different attributes of the visual field. Does this have anything to do with the fact some people have better eyesight than others or is that
mainly in the lens of the eye. You asked a problem which is at least as big as the one we've been talking about. Our eyesight might erupt in so many places in the lens in the retina in the optic nerve in the brain in the cortical areas of the brain in the perceptual areas of the brain. It's really quite impossible to say anything about that going about the basic results of your work very well. I think. If you if you wanted me to use some half a lifetime's work up in a nutshell as it were I would have to say something like this. And this starts off a little bit by being an apology for why I'm working on humans. There are some aspects of biological research which for ethical reasons you can't perform on an intact living human subject.
Typically it's the sort of research which involves surgical interference with a subject and the standard procedure in all of these cases then is to perform the experiments on animals if possible using animals as near to human as you can manage monkeys and then to extrapolate the results from. The animal to the human vets approach is and has turned out to be a very valuable one in the visual system where you can really only get reliable results by using implanted electrodes and this immediately puts you in the animal type of experiment. But extrapolating from one species to another is always prone to error especially if you're extrapolating to humans because humans have advanced so far beyond their nearest cousins.
So you can find many great actions of monkeys all close together but from the nearest monkey to the human is a big big jump. So although we perform these animal experiments we are always seeking experimental ways of confirming our guesses. That the functional organization of the human physical system now in our case implanted electrodes in animals have shown that in the retina and in the cortex abstractions are made of the visual field and that is long before the information has reached a perceptual level. The very earliest levels of neuro normal levels in the human or in these animals rather are starting to abstract the information now out of the potential measurements show changes of the of a potential which can be interpreted to mean that the human visual cortex performs exactly the same sort of function that the human is in
- About science
- About vision
- Producing Organization
- California Institute of Technology
- KPCC-FM (Radio station : Pasadena, Calif.)
- Contributing Organization
- University of Maryland (College Park, Maryland)
- AAPB ID
- This program focuses on the study of vision. The guest for this program is Dr. Derek Fender, California Institute of Technology.
- Interview series on variety of science-related subjects, produced by the California Institute of Technology. Features three Cal Tech faculty members: Dr. Peter Lissaman, Dr. Albert R. Hibbs, and Dr. Robert Meghreblian.
- Media type
Guest: Fender, Derek H.
Host: Hibbs, Albert R.
Producing Organization: California Institute of Technology
Producing Organization: KPCC-FM (Radio station : Pasadena, Calif.)
- AAPB Contributor Holdings
University of Maryland
Identifier: 66-40-80 (National Association of Educational Broadcasters)
Format: 1/4 inch audio tape
If you have a copy of this asset and would like us to add it to our catalog, please contact us.
- Chicago: “About science; About vision,” 1968-03-10, University of Maryland, American Archive of Public Broadcasting (GBH and the Library of Congress), Boston, MA and Washington, DC, accessed June 22, 2021, http://americanarchive.org/catalog/cpb-aacip-500-7m042m59.
- MLA: “About science; About vision.” 1968-03-10. University of Maryland, American Archive of Public Broadcasting (GBH and the Library of Congress), Boston, MA and Washington, DC. Web. June 22, 2021. <http://americanarchive.org/catalog/cpb-aacip-500-7m042m59>.
- APA: About science; About vision. Boston, MA: University of Maryland, American Archive of Public Broadcasting (GBH and the Library of Congress), Boston, MA and Washington, DC. Retrieved from http://americanarchive.org/catalog/cpb-aacip-500-7m042m59