Chapter 11 Chapter 9 Several Types of Experiments

"Research is an art, that is, the art of how to design some solutions to solve those problems." —Sir Peter Medwar Strictly speaking, all that each person can be sure of is that he is conscious.For example, I know that I am conscious.It seems to me that your behavior is very similar to mine, and especially that you have convinced me that you are conscious, so I can safely infer that you are also conscious.If I were interested in the nature of my own consciousness, I would not have to confine my research to myself, but could well experiment on others, provided they were not in a coma.

Psychological experiments on conscious subjects are not enough to uncover the neural mechanisms of consciousness.We also had to study the nerve cells, molecules and their interactions in the human brain.We can get most of the information about brain structure from the brains of the dead.But to study the complex behavior of nerve cells, experiments must be done in vivo.There were no insurmountable technical problems in the experiment itself.It is more due to ethical and moral considerations that make many such experiments impossible or very difficult. Most people are not opposed to having electrodes placed on their scalp to measure brain waves.But removing part of the skull, even temporarily, to insert electrodes directly into living brain tissue is unacceptable.Even if someone is willing to undergo a craniotomy experiment for scientific discoveries, no doctor will agree to perform such an operation.He would say it was a breach of his Hippocratic oath, or more likely someone would sue him for it.In our society, people will voluntarily join the army and do not hesitate to be injured or even sacrificed, but they may not be willing to accept those dangerous experiments just to obtain scientific knowledge.

A few intrepid researchers experimented on themselves.British biochemist and geneticist Haldane (J.B.S.Haldane) is a famous example.He even wrote an article on it called "On Being Ones Own Rabbit" (On Being Ones Own Rabbit), along with some of the most famous stories in the history of medicine, such as Sir Ronald Ross. Ronald Ross proved in himself that mosquitoes transmit malaria.But otherwise, serving as subjects for experiments that might help satisfy scientific curiosity is discouraged, even forbidden. In some cases, it is necessary to perform brain surgery on some patients while awake.Thus, if the patient consented, some very limited experiments could be performed on the bare brain.Since there are no pain receptors in the brain, patients do not experience discomfort from mild electrical stimulation of the exposed brain surface.Unfortunately, the time available for experimentation in surgery is often short, and few neurosurgeons are motivated by their interest in the subtle workings of the brain.This type of research was pioneered in the middle of this century by Canadian neurosurgeon Wilder Penfield.More recently, George Ojemann of the University of Washington School of Medicine in Seattle has led research in this area.He suppressed the activity of neurons in a small area near the electrodes with brief bursts of stimulating electrical current.If the current is weak enough, removing it won't cause permanent effects.He focused on areas of the cortex associated with language; this was because when he removed parts of the cortex from patients to reduce their likelihood of seizures, he wanted to damage as little of the adjacent language areas as possible.

Ogerman has a well-known experimental result.The patient has spoken English and Greek since childhood.When some areas on the surface of the neocortex on the left side of the brain were electrically stimulated, she was temporarily unable to use certain English words, but this did not affect her use of the corresponding Greek words. The opposite happened when other parts were stimulated, suggesting that both There are striking differences in where certain features of language are localized in the brain. For the most part, we can only study the behavior of the human brain from outside the skull. ① There are many different scanning methods to obtain images of living brains, but they all have great limitations in spatial or temporal resolution.Most methods are prohibitively expensive and limited in use due to medical considerations.

It's no wonder, then, that neuroscientists give priority to experimenting with animals.While I'm not convinced that a monkey is as conscious as you are, I have reason to think it's not quite an automaton, that is, a machine that behaves complexly but is completely devoid of awareness.This is not to say that monkeys have the same self-awareness as humans.Some experiments, such as the mirror-recognition experiment, suggest that certain apes, such as chimpanzees, may possess a degree of self-awareness.With monkeys, there is little, if any, self-awareness.But there is still reason to boldly assert that monkeys have a visual consciousness similar to that of humans, except that it cannot be expressed in words.

For example, a macaque can be trained to distinguish two colors that are very similar.These experiments show that the performance of macaques is comparable to that of our humans, within about 2 times.This is far from true for cats, which are mainly nocturnal, and even more so for rats.Because chimpanzees and gorillas are too expensive to do harmful experiments with them, if our main concern is the molecular characteristics of the mammalian brain, then rats and mice are the best and cheapest experimental animals.Although the features of their brains are in many ways simpler than those of ours, the molecules of their brains may be very similar to ours.

There is also the advantage of using monkeys and other mammals rather than humans in that they are currently better suited for neuroanatomical research.the reason is simple.Almost all modern approaches to the study of long-range connections in the brain make use of the active transport of molecules up and down in neurons.To do this, a chemical is injected into a certain part of the brain of a living animal.The substance is transported in the brain along the connections between neurons to other parts of the brain directly connected to the point of injection.This process usually takes several days.Thereafter, the experimental animals are euthanized and the sites where the injected substance has reached are examined.It is obviously impossible to do this kind of experiment with humans. Due to this limitation, our understanding of the long-range connections of the macaque brain is much richer than our understanding of our own.

One might think that this apparent gap in knowledge would worry neuroanatomists; since the human brain is not identical to that of a macaque, they would specifically demand new ways of studying human neuroanatomy, but this is not the case.In fact, now is the time to change our neuroanatomical defects in the human body, and those far-sighted foundations should immediately start inventing related new technologies. Even as we devise new ways to study neuroanatomy in humans, there are still many critical experiments that can only be performed in animals.These experiments sometimes last for several months.Although most experiments involve little or no pain, the animals often need to be killed (still painlessly) after the experiment is over.Animal rights groups are certainly right to insist that laboratory animals be treated well.Thanks to their efforts, lab animals are better cared for now than they used to be, but it would be too sentimental to idealize animals.Wild carnivores and herbivores often lead harsh lives and have shorter lifespans compared to the lives of captive animals.One view claims that since humans and animals are "parts of nature," they should be treated completely equally.It doesn't make sense.Should a gorilla really deserve a college education?The insistence on treating animals exactly like humans devalues ​​our uniquely human capabilities.Animals should be treated humanely, but to put them on an equal footing with humans is a distorted value.

What are the limitations of monkeys as experimental subjects in neuroanatomy and neurophysiology?It is possible to train clever monkeys to complete some simple psychological tests, but it is laborious.In one experiment, macaques were asked to maintain their gaze (that is, fixate on the same point).It presses one lever when it sees a horizontal line and another lever when it sees a vertical line.Such training usually takes several weeks or even longer.And how easy it is to let college graduates do this experiment!Furthermore, people as subjects can verbally describe what they see.They can also tell us what they have imagined or dreamed.It is almost impossible to get this kind of information from monkeys.

It appears that only one strategy is feasible.This is doing certain different types of experiments on humans and animals.This requires assuming how similar (and different) monkey brains are to human brains, but this is risky.No big progress can be made without risk.We should therefore be bold enough to approach research in this way, but cautious enough to check the validity of our assumptions as often as possible. One of the oldest methods of studying brain waves is the electroencephalogram (EEG).It places one or more coarse electrodes directly on the scalp.There is a wealth of electrical activity in the brain, but the electrical properties of the skull interfere with extraction of the electrical signals.A single electrode will pick up the electric field signal produced by as many as tens of millions of neurons, so the signal contributed by a single neuron to the electrode is drowned out by the activity of a large number of neurons in its neighbourhood.It's like trying to study the conversations of people in a city from 1,000 feet up.You can hear people yelling at a football stadium, but you can't tell what language people are talking there.

The biggest advantage of EEG is that the time resolution is quite high, roughly around 1 millisecond.This allows the rise and fall of brain waves to be recorded fairly well.It's not quite clear what these waves mean.There are very significant differences in the brain waves of the waking state and the slow-wave sleep state.Brain waves during REM sleep are similar to those of waking.Therefore, it has another name - abnormal sleep, that is, a person is in a state of sleep, but his brain seems to be awake.Most of our dreams occur during this stage of sleep. A common technique for recording brainwaves is to record immediately after some sensory input, such as a sharp click heard by the ear.Responses elicited by stimuli are usually small (i.e., low signal-to-noise ratio) compared to background electrical noise.Therefore, little can be seen from a single response and the experiment must be repeated many times with all signals averaged relative to the onset of each event.Because the noise is always averaged out, this improves the signal-to-noise ratio and often results in a perfectly repeatable profile of the typical brain waves associated with brain activity.For example, there is often a spike called P300 in the response, where P means positive potential, and 300 means that there is a time interval of 300 milliseconds between the stimulus signal and the spike (see Figure 35).It is usually associated with something startling and requiring attention.I'm guessing it's roughly a signal from the brainstem to the higher brain areas that remember the (stimulus) event. Unfortunately, it is difficult to pinpoint the location of neural activity that generates such event-related potentials.The problem is that if we know the electrical activity of each neuron, we can mathematically calculate the effect of placing electrodes anywhere on the scalp.Conversely, the electrical activity obtained from the electrodes cannot calculate the electrical activity in all parts of the brain.In theory, there are an almost infinite number of distributions of brain activity that could produce the same signal on the scalp.Still, even if it's impossible to recover all the details of neural activity, it's still hoped to gain some insight into where most of this activity occurs.By placing a certain number of electrodes across the scalp, we can get a good idea of ​​where most of the neural activity is located.If one electrode recorded a large signal and the others all had smaller signals, then most of the neural activity likely occurred near the electrode that recorded the large signal.Unfortunately, the situation in the experiment is much more complicated (1). Some limited but very useful information can be obtained from these event-related potentials.For example, the auditory part of the cortex is mainly located near the temporal lobe of the brain.What would it be like there if a person was born totally deaf?One study selected deaf people whose parents were also deaf.Thus it is almost certain that their birth defect is genetic, and the defect may lie in the structure of the ears rather than in the brain.Looking at event-related potentials, psychologist Helen Neville and her colleagues found that certain responses to signals in the periphery of the visual field in these patients had a much larger spike (with a delay of about 150ms).These enhancements were seen in parts of the anterior temporal and frontal lobes normally associated with hearing. This heightened response to signals from the periphery of the visual field is not surprising, because when deaf people sign to each other, their gaze is mainly fixed on the gesticulator's eyes and face.Therefore, most of the gesture information comes from the peripheral region of the gaze center.As a control, Neville also studied subjects whose parents were deaf but who were hearing and had learned American Sign Language.They did not experience the increased neural activity seen in the congenitally deaf subjects.This suggests that learning ASL does not induce the above-mentioned reinforcing effects. Neville speculates that part of the visual system somehow replaces part of the auditory system during brain development because completely deaf people lack normal sound-related neural activity.In people with hearing, it may be that normal auditory input prevents any visual areas from displacing the auditory areas of the cortex.Current animal experiments suggest that this idea makes sense. A more recent technique studies the changing magnetic fields produced by the brain.This magnetic field is extremely weak, only a tiny part of the Earth's magnetic field.Therefore, special detectors called squids (short for superting quantum interference devices) are used, and the changing magnetic fields in the environment are carefully shielded so that the whole device is not disturbed.Originally only one squid was used, but now a set of 37 of these probes is used.It is generally better spatially localized than EEG.Furthermore, it has similar advantages and limitations to electric fields, except that the skull interferes much less with magnetic signals.The magnetic probe responds to a source of dipoles perpendicular to the electrical dipoles that generate the EEG and thus detects the signal lost by the EEG.vice versa. While squids probes aren't cheap, conducting experiments to study brain waves isn't terribly expensive.Other major scanning methods not only require expensive instrumentation, but are also expensive to run.These scanning devices are few in number and almost all are owned by medical institutions.They can only produce moving images of one slice of the brain at a time.Thus, imaging of several slices is usually required to cover a region of interest. Broadly speaking, there are two types of scanning techniques, which detect the static structure and dynamic activity of the brain respectively.One of the earliest techniques, called CAT scanning, or computer-aided tomography, used modulation rays, and a more modern technique, magnetic resonance imaging (MRI), produces excellent high-resolution images.As far as is known, it does not cause brain damage to the experimenter.In normal use, it records the density of protons (i.e. hydrogen nuclei) and is therefore particularly sensitive to water.The resulting image has good contrast, but the image is static and does not record brain activity (see Figure 36).Beyond that, both methods clearly show broad structural differences between different brains.Under the appropriate circumstances, both methods can detect structural damage caused by blows, gunshot wounds and other injuries to the brain.It's just that the types of damage that are easily detectable by different technologies are different.Using a special technique, MRI scans can produce a three-dimensional reconstruction of a living human brain, including its appearance.Figure 37 is a profile of the brain of neurophilosopher Patricia Churchland. Positron emission tomography (PET) is a different approach.It records local brain activity, but averages that activity over a period of about a minute or so.The experimenter is injected with a chemical, usually water, labeled with a harmless radioactive atom (such as 15O).The radioactive atom emits a positron when it decays. ① The marked water enters the blood. 15O has a very short half-life, which means that it must be produced within a short period of time from cyclotron production to injection into the human body.But it has two advantages: Oxygen decays very quickly, so a second experiment can be done about ten minutes later; and radioactive materials are very short-lived, which means that the total radiation dose to the experimenter in order to obtain the desired signal is very high. The damage caused is negligible.Thus the method can be used in healthy volunteers and not necessarily limited to frail patients. When a part of the brain has more neural activity than usual, the blood supply to it also increases.In effect, the computer-generated images corresponded to scans of blood flow levels in various parts of the brain.Other experiments scanned subjects in a control condition.The difference between the two maps is roughly consistent with changes in neural activity when the brain is in the stimulated and controlled states. This technique has yielded a number of interesting and challenging results.Of particular note is a research group led by Marcus Raichle at Washington University School of Medicine in St. Louis.In earlier experiments they studied responses to a small set of visual patterns.These patterns were selected to produce maximal responses in distinct, rather broad areas of the cortex.Changes in blood flow in primary visual areas of the neocortex were much the same as those predicted by earlier studies of injury in the human brain.Changes in blood flow to other visual areas of the cortex were also found, but whether they were of value was unclear. They recently studied changes in blood flow during what is known as the "Stroop interference effect."This is a more complex vision task.In the experiment, the subjects were asked to recognize the color of a word as quickly as possible.For example, the captured target might be the word red printed in green.The difference between the color of the word (green) and the meaning of the word (red) caused the subjects to increase their reaction time.Comparing the distribution of blood flow in this task with another direct situation (where the word red was printed in red), they found that several cortical areas showed blood flow in the Stroop condition. The phenomenon of increased flow, in which the region with the largest increase is the "right anterior cingulate gyrus", which is in the middle of the brain, near the frontal area.They think it has to do with the level of attention required to complete the task.From this they concluded: "These data suggest that the anterior cingulate gyrus is involved in a selective process of competing between the Sexual alternation." I feel this statement is closer to what we think of as free will than to attention in the usual sense (see the postscript at the end of this book).Clearly, we need to learn more about the neural mechanisms involved in the different processes. PET scanning can achieve some results that are difficult to obtain by other methods, but it also has several limitations.In addition to being expensive, its spatial resolution is not very high (although it has gradually improved with most modern instruments), currently usually around 8mm.Another downside of it is that the time resolution is rather poor.It takes about a minute to get a good signal, whereas EEG works in the millisecond range. Some leading research centers currently use PET scans in combination with MRI scans. PET records brain activity, while MRI obtains brain structure, so PET scans can be mapped onto the same person's brain, rather than onto an "average" brain as has been done in the past.However, it is not long before the interpretation of these results encounters the aforementioned limitations arising from a lack of detailed neuroanatomical knowledge. Now some new methods using MRI scans have been developed.One of these methods is particularly sensitive to lipid compounds.Scanned images can be used to help locate a number of different cortical areas in a person (the exact location of these areas varies from person to person).This is due to the fact that some cortices have more myelinated axons than others, containing more lipids. Other new MRl methods attempt to probe all kinds of metabolic and other brain activity, not just its static structure, but they all appear to have lower signal-to-noise ratios than conventional MRl.It is thus expected to see the development of these new methods. The research on the human brain will be described here first.Is there any way to observe the behavior of neurons in animal brains?One way is to use thinner electrodes to get the most detailed information.This is an insulated wire with the tip exposed.After the animal is anesthetized, part of the skull is removed and electrodes are placed just inside the neural tissue.Since there are no pain receptors in the brain, the electrodes do not cause pain to the animals.As long as the tip of the microelectrode is very close to a cell, it can detect when it fires outside the cell.It can also pick up weaker signals from farther away cells.By moving the tip of the electrode along its length through the tissue, the activity of nerve cells can be detected one by one. The experimenter can choose where to place the electrode in the animal's brain, but in a sense what exactly he is recording Which type of cell it is depends entirely on luck.Now people often use a set of electrodes for recording, so that the activity of more than one neuron can be detected at the same time. Another technique is to study a very thin slice of neural tissue taken from an animal's brain.The electrode used here is a very small glass tube with a tapered tip.Carefully place the electrode so that its tip is just inside a nerve cell.This yields more detailed information about the activity of that neuron. (This technique can also be used in anesthetized animals without damaging the brain, but it is much more difficult to use in conscious animals.) Brain slices can last for many hours if soaked in a suitable culture medium.Brain slices are easily perfused with different chemicals to examine their effects on neuronal behavior. Under certain conditions, neurons taken from the brains of very young animals were able to grow in the dish and spread out in all directions.Such neurons grow in contact with neighboring neurons, conditions that are farther away from the environment of a living animal, but which can be used to study the fundamental behavior of neurons' internal connections.These connected membranes have channels.When the channels are open, charged atoms (i.e. ions) are allowed to flow through. Perhaps most surprising is the current possibility to study the behavior of individual molecules in individual ion channels.This is achieved by a technique called "patch clamp".Erwin Neher and Bert Sakinann were awarded the 1991 Nobel Prize for developing and applying this technique, using a small glass pipette with a special The slanted tip, approximately 12 microns in diameter, is capable of aspiration of a small piece of the lipid membrane.With any luck, at least one ion channel will be included in the patch.The current flow through the membrane can be studied via electrical amplifiers and recording devices.The concentrations of the relevant ions remain at different values ​​on both sides of the small membrane.When the channel opens, even for a brief moment, a flood of charged ions rushes through it.This surging tide of ions generates a measurable electrical current.This is true even if only one channel is open.This allows one to study the effects of neurotransmitters and other pharmaceutical agents (usually other small organic molecules), as well as the effect of membrane voltage. Patch clamp was also used for another study on ion channels.The gene for this channel was artificially introduced into unfertilized frog eggs.Under the guidance of these foreign genes, the oocyte (that is, the unfertilized egg) synthesizes the protein of this channel and places it in the outer membrane.In this way, it can be sucked out with patch clamp.This technique is useful for discovering the gene for a particular ion channel. Now to summarize, there are many ways to study the human and animal brains.Some of these methods work from the outside of the skull, while others go directly to the inside of the brain.All methods have limitations in one way or another, either insufficient temporal or spatial resolution, or are prohibitively expensive.Some results are very easy to interpret but provide fairly limited information; other measurements are easy to make but difficult to interpret.Only by combining different approaches can we hope to unravel the mysteries of the brain. ①In rare cases, permanent electrodes must be implanted very deep in the brain tissue for medical reasons.However, the number of implanted electrodes is very small, so the information that can be obtained is also very limited. ①A commonly used approximation is to assume that there are four centers in the brain that generate most of this electrical activity.In this way, it is possible to find the approximate location of these centers by mathematical means.One way to test the validity of this assumption is to assume the existence of five centers and repeat the above calculation.If the resulting four centers are strong and one is very weak, then the four center approximation can be quite effective.Even so, this is just an educated guess. ① A positron travels a short distance before combining with an electron.After combining, both particles are annihilated, and their mass is converted into radiation, as two rays moving in almost opposite directions.Recording these gamma particles is a ring-shaped coherent counter.A computer combines all the decay traces and analyzes the regions most likely to have produced these gamma rays.