Jump to ContentJump to Main Navigation
Matter of MindA Neurologist's View of Brain-Behavior Relationships$

Kenneth M. Heilman

Print publication date: 2002

Print ISBN-13: 9780195144901

Published to Oxford Scholarship Online: May 2009

DOI: 10.1093/acprof:oso/9780195144901.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 17 September 2019

Attention

Attention

Chapter:
(p.87) CHAPTER 4 ATTENTION
Source:
Matter of Mind
Author(s):

Kenneth M. Heilman

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780195144901.003.0004

Abstract and Keywords

This chapter discusses how we become aware of stimuli, focus our attention, and withdraw it. Topics covered include sensory awareness, spotlights and floodlights of attention, and habituation. It shows that the brain has a limited processing capacity and receives more information than it can process fully. People attend to stimuli that are important and ignore those that are unimportant. Stimulus significance is determined by their immediate needs (drives) and future goals.

Keywords:   sensory awareness, habituation, spatial neglect, neglect, unilateral neglect, hemispatial neglect, extinction to simultaneous stimuli, line bisection test, cancellation test

William James, a Harvard-trained physician and one of the founders of American psychology, said that attention is one of those terms that is difficult to define, though everybody knows what it means. This term may be difficult to define because, rather than being an object, it is a process that is not well understood. Humans need to use attentional processing because our brains have a limited capacity. That is, they receive more information than they can process simultaneously and fully. While you are reading this book, if you find it interesting, you are attending to the words on the page and are unaware of how your left foot feels—until I mention your left foot. Now you can shift your attention to it and feel it. In this chapter we will discuss how we become aware of stimuli, focus our attention, and withdraw it.

(p.88) SENSORY AWARENESS

Clinical Observations of Unawareness

When I was a medical student at the University of Virginia, there was a hierarchy, and the medical students were at the bottom. To make it easier for the house staff and faculty to recognize us, we wore short white coats with normally colored pants such as chinos. The interns, residents, and fellows wore short white coats with white pants. It was easy, however, to tell the interns from the residents and fellows. They always looked tired, and their whites never seemed to be as white as the residents’ uniforms. The attending physicians, who were the faculty of the medical school, wore long, white, neatly pressed coats. The dress code allowed the medical staff and patients to know that we were students.

On the way to picking up a bottle of urine from a patient, I passed a middle-aged man who was sitting on the edge of his bed eating his lunch. He said, “Doc, can you come over here?” I looked around to see if any of the house staff or attending physicians were nearby, but I was the only one on the ward at that time, so I walked over to the man's bed and said, “I am not a doctor yet, but I would be happy to try and help you.” He said, “Being a medical student is good enough for me to call you ‘Doc’ “ When I thanked him and offered to help, he asked, “What kind of weird place is this? Look at my food tray. They only served me vegetables. I am not on a special diet. How do I get some meat?” I looked down at his tray. On it was a hospital plate with a divider to separate the ground-up meat from the ground-up vegetables. The vegetables were on the right side of the plate and the meat, which didn’t look quite edible, was on the left side. On rounds, I had heard that this man had been admitted after a stroke. I thought that perhaps the stroke had injured the right occipital lobe, containing the visual cortex, which is important for seeing the left side. Because I thought that the patient could not see on the left side, I rotated the dish 180 degrees so that meat was now on the right side. He said, “Thank you, Doctor, for finding me some meat.”

When this man finished eating, I came back to examine him. Blindness on one side is called hemianopia (loss of half vision). This loss of vision usually occurs in both eyes. To test the patient for hemianopia, I stood facing him and asked him to look at my nose.

(p.89) I then told him that I would move either my right or my left hand, and he was to tell me which hand I moved. When I moved my right hand, which normal people see with their right occipital cortex, he had no problem seeing my moving hand. Since his left-sided vision was normal, I did not understand why he could not see the left side of his plate. The other main finding of my examination was severe weakness of the patient's left arm, which he kept in a flexed posture on his chest.

The attending physician that month was a New Zealander named Dr. Fritz Dreyfus, a superb neurologist and teacher. The next day, at the end of rounds, when he asked if anyone had questions, I asked him to explain why this patient, who was not blind on his left side, could not see the meat on the left side of his plate. Even when physicians cannot explain symptoms, they can name them. For example, some people might develop red bumps on their skin. The dermatologist who examines them may not know what is causing the rash but will tell them that they have erythema nodosum, which in Latin means red bumps. Dr. Dreyfus said that the man had what is called unilateral neglect. He added that Derek Denny-Brown, the most famous neurologist from New Zealand, had written about this disorder, and he suggested that I read his articles.

After we finished rounds, I went to the library and started reading everything I could find about unilateral neglect. I learned that when patients with this condition try to draw or copy a figure, they often leave out the side of the picture opposite the side of their brain injury (contralesional). In another test for unilateral neglect, patients are shown a horizontal line and asked to locate its center. The patients with neglect respond by deviating their mark to the non-neglected side of the line, an ipsilateral spatial bias. Lastly, there is the cancellation test, developed by Simon Horenstein and later reported by Martin Albert to be a sensitive indicator of neglect. In this test, the physician draws many small lines, about 1 or 2 inches long, and distributes them randomly over a page of white paper. He or she then shows the patients this paper and asks them to cross out, or cancel, all of the lines. Patients with neglect fail to cross out the lines on the side of the page opposite to their brain lesion.

I went back to the wards to see the man with unilateral neglect. After speaking with him for several minutes about the reasons he came to the hospital, I was surprised that he appeared so calm and indifferent when his left side was so weak. His name was George Jones, (p.90) and he was a 67-year-old retired civil engineer. I asked him if I could perform some tests and he replied, “Yes, you are the doctor.” I also received approval from the medical team that was taking care of him.

The first test I gave him was the line bisection test. I drew a line about 10 inches long on a blank piece of paper, put it in front of him, gave him a pencil, and asked him to put a mark in the middle of the line. He said, “Doc, you mean bisect the line.” I nodded. He said, “Doc, you probably remember from geometry that the way you bisect a line is to draw either an equilateral or isosceles triangle and then drop a meridian.” He then proceeded to draw an isosceles triangle. The two sides of the triangle were equal, as they should be. But whereas the right leg of the triangle did come down to the right end of the horizontal line, the left leg of the triangle came down to the middle of the line rather than to the left end. Therefore, when the patient dropped his meridian, it intersected the horizontal line several inches to the right of center. He then darkened this intersection of the horizontal line and his meridian, pointed to this intersection with the index finger of his right hand, and said, “There, the line has been bisected” (Fig. 4–1). Considering how he missed seeing the meat on the left side of the plate, his performance on this line bisection task was not surprising. Mr. Jones's performance illustrates that if you initially misperceive the world, even if you perform formal operations your behavior remains deviant.

Next, I drew a simple picture of a daisy and asked him to copy

                   Attention

Figure 4–1. A reproduction of Mr. Jones's attempt to bisect the line by first drawing an isosceles triangle and then dropping a meridian. However, because the left end of the triangle was not on the left end of the line, his meridian is not at the midpoint of the line.

(p.91)
                   Attention

Figure 4–2. A reproduction of Mr. Jones's attempt to copy the picture of the daisy that I had drawn.

it. His drawing showed neglect (see Fig. 4–2). He did not draw petals on the left side of the flower, and he placed the drawing on the far right side of the page. Lastly, I gave Mr. Jones the cancellation test; he canceled only the lines on the right side of the page (Fig. 4–3).

Brain Mechanisms

It was apparent that Mr. Jones was unaware of stimuli on the left side. But why? That weekend, admissions to the hospital slowed down and I was able to read more about this unilateral, or hemispatial, neglect. Denny-Brown wrote that after somatosensory (touch, pain, heat/cold, position), visual, and auditory stimuli come to the cerebral cortex, this sensory information converges in the

                   Attention

Figure 4–3. A reproduction of a cancellation test. Mr. Jones was asked to cancel all the lines. He failed to cancel the lines on the left side of the page.

(p.92) parietal lobes. The senses that monitor the left half of space meet in the right parietal lobe, and those that monitor the right half of space meet in the left parietal lobe. Denny-Brown suggested that this multimodal sensory synthesis allows us to develop awareness of the right and left halves of space.

While I was in medical school, I had no further opportunities to learn about neglect. After training in internal medicine, I decided to become a neurologist. I took my residency at the Harvard Neurological Unit of Boston City Hospital because that was where Denny-Brown was Chairman. During my residency, I was asked to see a patient on the medical service. The patient was brought to the hospital by his family because he had sudden onset of left-sided weakness, or hemiplegia. After performing a physical examination, I tested this patient using some of the same tests I had used with Mr. Jones. On the line bisection task, he thought that the midline was 2 inches to the right of the actual midline. On the line cancellation test, he canceled only the lines on the right half of the page. Like Mr. Jones, he was unaware of stimuli on the left side of space. When I asked why he came to the hospital, he replied, “Because my family brought me.” I diagnosed this patient as having had a stroke of the right parietal lobe. Shortly after completing my evaluation and writing a note on his chart, I received a call from the medical resident caring for this patient. He asked if I wanted to obtain a radioisotope scan of his brain. Compared to today's brain imaging methods this radioisotope scanning was not very revealing, and I felt confident that the patient had had a stroke that injured his right parietal lobe. In addition, radioisotope scanning was not available in Boston City Hospital. Therefore, I, told the medical resident that I was pretty confident about my diagnosis and did not think there was any need for a radioisotope scan.

Two days later, I again saw this medical resident, who was carrying an x-ray film. He caught up with me and said, “Ken, remember my patient with left-sided neglect whom you saw a couple of days ago?” I nodded. He said, “Well guess what?” This question was cause for panic. Before CT imaging was available, neurologists were frequendy concerned that by relying on the physical history and examination alone, they might have missed a tumor or blood clot that could be removed surgically. The main way to find out if a patient had such a lesion was to perform a test called an angiogram or arteriogram. It involved injecting a dye directly into the carotid artery in (p.93) the neck, which carries blood to the brain, and then obtaining a series of x-rays of the skull. The dye allowed us to see the arteries. In addition to seeing if an artery was blocked, we could see if the arteries were being displaced by a tumor or blood clot. Although the arteriogram was a good diagnostic test, about 3%-5% of the patients who were given it in our hospital had a serious side effect such as a stroke. Thus, we did not use this test if we thought the patient had already had a stroke, but only when we believed there was a strong possibility of a tumor or a blood clot on the surface of the brain. I asked the resident if he had the results of an arteriogram on the x-rays. He said, “No, not an arteriogram; it is a radioisotope scan.” He then gave it to me. Rather than showing a stroke that had damaged the parietal lobe, the scan revealed damage to the frontal lobe. I said, “I never knew that patients could get neglect from damage to their frontal lobe. I need to review the literature to see what I can find. Can you get me a copy of this scan?” The resident said, “This is your copy. I thought you would want one.” I said, “Thanks! How did you get this scan?” He answered, “Don’t ask.”

The next weekend, I was on call. This meant staying and sleeping in the hospital for more than 48 hours. All the patients on our inpatient service were stable. There were only a few consultations and calls from the emergency room, so I could spend a lot of time in the library. One of the modern inventions for which I am most thankful is the Medline service. If you want to review all the articles written on neglect, all you have to do is log on to Medline and type the words neglect and frontal lobe. The computer will search the National Library of Medicine, and within seconds will display all the articles written on this subject since the 1960s. Although the full articles are not always available, you can get an abstract. If the abstract looks interesting, you can read the journal in the library. In the 1960s, however, it was necessary to search through a set of thick, heavy books called Index Medicus. Each year a new volume appeared. I had to go through ten of these books but found nothing about the frontal lobes and neglect. In general, there was also very little about neglect alone.

By the time I had seen this man with frontal lobe neglect, Denny-Brown had retired as Chairman of Boston City Hospital's Harvard Neurological Unit and Norman Geschwind, who had been Chairman at Boston University, had taken over his position. One of the nice customs Dr. Geschwind started was to have morning coffee (p.94) in a small meeting room with several chairs around a table. During these times, a resident or faculty member could sit down and chat with him about anything that was on his or her mind. The day after seeing the scan on this man with the frontal stroke I attended this morning coffee. A good observer, Dr. Geschwind asked, “What's on your mind, Ken?” I told him about the patient with neglect having a frontal lesion and showed him the brain scan. He said, “Interesting. In almost all the reports of experimental neglect in monkeys, the neglect was induced by making frontal lobe lesions.” Not wanting to spend another day with Index Medicus I asked, “Do you recall the reference?” He said, “Welch and Stuteville wrote a paper published in Brain about 10 to 15 years ago”. Our departmental library had many older issues of Brain, and I had no difficulty finding the article. I made one copy for the medical resident and one copy for myself.

In most neurology journal articles, the author reviews the past literature before describing his or her new observations. Two things struck me about this article. Although previous investigators as far back as the late nineteenth century had reported neglect-like behavior in animals with frontal lesions, there was no mention that humans develop neglect if they have injuries in the same regions of the frontal lobes. In addition, the anatomical location of a posterior lesion that could induce neglect in monkeys was not entirely elucidated.

After Norman Geschwind came to Harvard, he recruited Deepak Pandya, a neuroanatomist who is interested in how different parts of the cerebral cortex are connected anatomically. Studying monkeys’ brains, Pandya found that each of the primary sensory reception areas in the cerebral cortex areas projects only to its own association areas. Thus, the primary visual area is connected to visual association areas, the primary auditory area projects to the auditory association areas, and the primary somatosensory area (e.g., touch) projects to the somatosensory association areas (Fig. 4–4). Each of these sensory- or modality-specific association areas projects to multimodal (polymodal) or supramodal sensory areas. In humans, one of these polymodal sensory areas is located in the inferior portion of the parietal lobe (Fig. 4–4). Unlike the primary sensory reception areas, which perform analyses of stimuli, the modality specific sensory association areas perform syntheses, and these syntheses allow the brain to form percepts. The supramodal areas receive input about percepts from several modality-specific association areas, as (p.95)

                   Attention

Figure 4–4. Diagram of cortical attention network. Visual, tactile, and auditory stimuli project to their primary sensory cortices, which are receiving areas that perform sensory analysis (visual primary cortex = VI, auditory = Al, tactile = S1. These primary cortical areas project to a higher level of cortex that helps synthesize sensory inputs and match them to stored memories, or representations, of previous experience in this modality (S2 = somatosensory association cortex, A2 = auditory association cortex, VTL + SPL = visual association cortex). Each of these sensory association areas then projects to a polymodal or supramodal area in the inferior parietal lobe (IPL), where sensory percepts activate higher concepts. This figure also demonstrates the strong connections between the frontal and parietal lobes (SPL = superior parietal lobe, VTL = ventral temporal lobe, DFL = dorsolateral frontal lobe).

well as input from other portions of the brain not directly related to sensory input. These supramodal areas allow people to make cross-modal associations (e.g., the shape of a dog and a bark) and help them determine the meaning of stimuli.

Primate brains are convoluted. A convoluted brain has mountains called gyri, valleys called sulci, and gorges called fissures. When I first visited my wife's family in West Virginia, I commented that West Virginia was a small state. My wife, who is proud of her home state, told me that if West Virginia was flattened, it might be larger than states without mountains such as Florida. That is, although West Virginia occupies a smaller space than Florida because it has so many mountains and valleys, its surface area is very large. The cerebral cortex of primates contains the neuronal systems that are critical for mediating complex activities. Therefore, as primate brains evolved, they needed more neuronal systems. An increase in neuronal systems allows greater intelligence by increasing the (p.96) storage capacity of the brain and the ability to make computations. To increase the surface area, where the neurons are found, without greatly increasing the size of the head, the brain developed gyri, sulci, and fissures. The banks of the sulci are like the sides of the mountains.

Deepak Pandya demonstrated that in monkeys the polymodal areas, where all the senses come together, are located in the inferior parietal gyrus and in both banks of the superior temporal sulcus. The superior bank of the superior temporal sulcus is the superior temporal gyrus, and the inferior bank is the middle temporal gyrus. Reading Pandya's papers reminded me of the paper on neglect written by Denny-Brown, who suggested that all the sensory modalities come together in the parietal lobes of humans and that this synthesis allows us to be aware of stimuli in the opposite half of space. Since in monkeys all the senses come together in the banks of the superior temporal sulcus and the inferior parietal gyrus, if these regions are destroyed in one hemisphere (e.g., the right), the animal should neglect stimuli that are presented on the opposite (left) side of space.

To test this hypothesis, I asked Dr. Pandya if we could destroy this polymodal area, on one side of the brain, in a few rhesus monkeys and see if this injury induced neglect. In other (control) monkeys we could destroy an equal-sized area in a different part of the brain that is not polymodal. He agreed, and the experiments were performed. The monkeys that had the parietal-temporal lesions demonstrated unilateral neglect, and the control monkeys did not. These observations appeared to support Denny-Brown's hypothesis. When examined these monkeys, however, something troubled me. If a stimulus was presented to the side opposite their temporal-parietal lesions, the monkeys sometimes appeared to be aware of this stimulus. However, if both the right and left sides were stimulated at the same time, the animal only appeared to be aware of the stimulus on the same side of the body as the cerebral lesion (ipsiateral side). According to Denny-Brown's hypothesis, these temporal-parietal lesions should have destroyed the representation of contralateral space. If so, why could these monkeys sometimes detect a single stimulus applied to the side opposite the lesion, but, when given simultaneous stimuli on both sides, fail to detect the stimulus on the side opposite their hemispheric lesion? Denny-Brown's hypothesis could not entirely account for these observations.

(p.97) This phenomenon of failing to sense a stimulus on one side when stimulated on both sides had been reported in humans by Morris Bender, a New York neurologist who practiced at Mount Sinai Hospital more than a half century ago. Dr. Bender called it extinction to simultaneous stimulation but did not speculate about the underlying mechanisms. However, Walther Poppelreuter an early-twentieth entury German neurologist, thought that unawareness of stimuli opposite a hemispheric lesion may be related to a defect in attention, but what is attention?

After finishing my internal medicine training and before beginning my neurology residency, I joined the Air Force. At this time, in the mid-1960s, I had to go to Alabama for basic training. Part of the training involved how to deal with a disaster when there are more injured people than can be cared for simultaneously with limited personal and supplies (limited capacity). In this situation, one doctor has to be the triage officer. As this officer examines the injured people, he or she categorizes them into four groups. Minimal means that the injuries are so minor that they will probably heal by themselves, without medical or surgical intervention. Immediate means that if the injured person does not get immediate attention, he or she will probably die or be permanently disabled. These are the people who must be treated first. Expectant means that the person is so badly injured that he or she will probably die despite intervention, or that caring for this person will be such a drain on the available resources that two or more people classified as immediate may die. Delay means that the person needs treatment but will still survive, with no or little disability, if treatment is delayed. Attention is, in part, a mental triage process. We attend to those external or internal stimuli that are most important to us. The significance of a stimulus is determined by several factors. We almost always attend to a novel stimulus because we have not yet determined its meaning. We also attend to stimuli that are important to us, as determined by our immediate needs (drives) and future goals.

When reading this book attending to your left foot is not important to you, so you are probably not aware of it, but if a bug crawled onto your foot you would immediately attend to your foot because you detected a new stimulus. If you had pain in this foot, you would attend to it because pain induces immediate needs or drives. If you were waiting for a salesperson to put on a new shoe, you would also attend to it because of future goals or motivations.

(p.98) According to the Russian physiological psychologist Y.N. Sokolov, the sensory cortices and sensory association areas store memories of incoming stimuli. When a stimulus is new, the person develops an orienting response and attends to the stimulus. According to our model, the information from sensory areas projects to the supramodal temporal-parietal region, which is critical in directing attention. In addition to receiving projections from the sensory association cortex, the temporal-parietal region receives strong anatomical projections from the dorsolateral frontal lobe. The temporal-parietal area also projects back to the frontal lobes (Fig. 4–4). The Russian neurologist A.R. Luria demonstrated that patients with frontal lobe injuries often lose their goal-oriented behavior. This observation implies that when the dorsolateral frontal lobe is injured, it cannot supply information about future goals and motives to the temporal-parietal region. In the absence of this information, people may not correctly perform triage or incoming sensory information and may not attend to important stimuli.

A phylogenetically older part of the brain, found in all mammals, called the limbic system, may also be important in the triage process. Whereas the sensory areas of the brain in the temporal, parietal, and occipital lobes monitor the external world, through hearing, touch, and vision, the limbic system monitors the inner world of the body. The limbic system is also important in mediating emotions such as fear. The portions of the brain that comprise the limbic system are widely distributed in the brain. To influence behavior, this system has to communicate with the other neural systems. The part of the limbic system located in the middle of each hemisphere is called the cingulate gyrus (Fig. 4–5). It has strong connections with the dorsolateral frontal lobes and the temporal and parietal lobes. Patients with terminal cancer who are in severe pain sometimes have an operation in which the neurosurgeon removes part of the cingulate gyrus. Afterward, when these patients are asked if they still have pain, they often answer, “Yes, but I do not pay as much attention to the pain as I did before surgery.”

The cingulate gyrus may also be injured by stroke and tumors. When Ed Valenstein and I were searching for patients who had neglect due to frontal lesions, we found several patients with neglect from lesions that had injured the cingulate gyrus. However, naturally occurring lesions often involve more than one structure. To determine if the cingulate gyrus was the area critical for inducing neglect, we removed it surgically from one side of monkeys’ brains (p.99)

                   Attention

Figure 4–5. Diagram of a midsagittal section in which the brain is cut in half from front to back. This diagram shows the corpus callosum, the major connection between the right and left hemispheres. Above the corpus callosum is the cingulate gyrus.

and found that these monkeys did neglect stimuli on the opposite side of space.

Barbara Haws, our chief technician, not only took excellent care of our monkeys, but also spent so much time caring for them that she was able to make several interesting observations. Our monkeys had large homes that included a fenced-in but open area. Rural north Florida has many snakes. While some are poisonous, most are harmless. The fences, however, prevented the snakes from entering the monkeys’ homes, and none were bitten. Yet Barbara noticed that when these monkeys were in the open area and saw a snake beyond the fence, they panicked. Many of our monkeys are born in captivity. Unlike people, who have language, monkeys cannot communicate verbally. Therefore, their mothers cannot tell them, “Watch out for snakes because, if they bite, you can get sick or even die.” We are uncertain how these monkeys know that snakes may be dangerous. This information may be inherited rather than learned, and much inherited information that is associated with fear (fight or flight) is stored in the limbic system. After removing the cingulate gyrus on one side of a monkey's brain, my colleague Bob Watson brought a plastic snake to the laboratory. When he stood on the same side of the monkey as the injured hemisphere (nonneglected side) and wiggled the snake, the monkey panicked, but when he stood on the side of the monkey opposite the injured cingulate gyrus and wiggled the snake, the monkey showed no signs of fear. It did not panic because the cingulate gyrus, which is part of the limbic system, was injured on one side, and this injury caused the monkey to neglect the snake.

(p.100) In both monkeys and humans, neglect can be caused by injury to the dorsolateral frontal lobe, die cingulate gyrus, or die inferior parietal lobe. These diree areas are highly interconnected and to-gedier form a network that mediates spatially directed attention. Based on the information it receives, the parietal lobe acts like a triage officer. During a disaster, after die triage officer decides if an injured victim is immediate, delayed, expectant, or minimal, he or she must label the victim so that the treating physicians and surgeons can work on or process those labeled “immediate.” The human brain separates attended from unattended stimuli by knowing the type of stimulus (what) and its location in die environment (where). To allocate attention correcdy, the human inferior parietal lobe must receive this “what” and “where” information. Studies of patients with injuries to the superior portion of the parietal lobe (Fig. 4–6) reveal that they have trouble locating objects in space. For example, they may get lost when using a well-known route, or when diey try to grasp or point to an object, tiiey may miss it. These patients, however, can recognize objects. These observations suggest that the superior portion of the parietal lobe is important in determining the spatial location of objects (the “where” system) but is not important for recognizing objects (the “what” system).

In contrast to injuries of the superior parietal lobe, bilateral injuries or lesions in die inferior portion of the occipital and temporal lobes (Fig. 4–7) produce a condition call visual object agnosia (a, without; gnosis, knowledge). These individuals may be able to see

                   Attention

Figure 4–6. Diagram of the superior parietal lobe (SPL), which is dorsal to the inferior parietal lobe (IPL).

(p.101)
                   Attention

Figure 4–7. Diagram of the ventral view of the brain demonstrating the location of brain injuries associated with visual agnosia.

clearly, but they cannot recognize objects or people. At the National Institutes of Health, Leslie Ungerleider and Mortimer Mishkin studied the visual system of monkeys and observed that after nerve impulses from visual stimuli enter the primary visual cortex, in the occipital lobe, they are analyzed and processed by two visual streams. One stream goes to the parietal lobe and is called the dorsal stream. The other goes to the ventral temporal lobe and is called the ventral stream (Fig. 4–8). As observed in humans, the ventral stream, the “what” system, is important in recognizing objects, and the dorsal stream, the “where” system, is important in determining spatial location. In monkeys, the “what” and “where” systems converge in the banks of the posterior portion of the superior temporal gyrus. This region also receives input from the dorsolateral frontal lobe (goal-oriented system) and the cingulate gyrus (motivational system). Many scientists believe that the banks of the monkey's posterior superior temporal sulcus evolved into human's inferior parietal lobe (Fig. 4–9). When our group of researchers in Gainesville destroyed both banks of the posterior portion of the superior temporal gyrus, they produced the same type of neglect that is seen in humans with inferior parietal lesions.

When a person attends to an important stimulus, he or she becomes more alert or aroused. When we were doing research on (p.102)

                   Attention

Figure 4–8. Diagram of the lateral view of the brain demonstrating the primary visual area (VI) branching into the ventral visual stream in the ventral temporal lobe (VTL), and the dorsal visual stream in the superior parietal lobe (SPL). Whereas the ventral stream mediates the “what” system, which is important in recognizing objects and people, die dorsal stream mediates die “where” system, which is important in recognizing spatial location.

the brain mechanisms of attention and neglect, Bob Watson and I saw a patient who had severe neglect of the left side of space. This deficit occurred suddenly, suggesting that the man had suffered a stroke. Because the neglect was severe, we thought the stroke had destroyed both the dorsolateral frontal lobe and the inferior parietal lobe of the right hemisphere. A CT scan of his brain, however, revealed no injury to either the frontal or the parietal lobe. Even the cerebral cortex was intact. Instead, he had a hemorrhage deep in the brain (Fig. 4–10), on the right side, in an area called the reticular activating system. Such hemorrhages are seen in patients with a long history of hypertension. As we mentioned in Chapter 3, about 50 years ago, Moruzzi and Magoun studied the reticular activating system in cats that were made sleepy by drugs. Using electrodes, they stimulated neurons in this activating system. On stimulation, the sleepy cats became aroused and alert. Based on these and subsequent studies, neuroscientists believe that the reticular activating system is important in arousing the brain. Physicians have learned that in people with diseases such as kidney or liver failure, toxins can accumulate and poison this area of the brain. Brain swelling can compress the reticular activating system, and drugs may poison it. Patients with these diseases first become inattentive, and as the toxin or swelling builds up, they grow less alert or even sleepy. Doctors (p.103)
                   Attention

Figure 4–9. Diagrams of the lateral view of the brain in the human and the old world (rhesus) monkey. In the human brain, the inferior parietal lobe (IPL) contains two evolutionary new gyri (mounds of brain tissue) which are not found in the monkey's brain: the supramarginal gyrus (40) and the angular gyrus (39). In the monkey's brain, the IPL contains a pattern of nerve cells that the anatomist Brodmann classified as Area 7. In the human, this same area is now located in the superior parietal lobe (SPL) because of the growth of the supramarginal and angular gyri. These new gryi probably evolved from die banks of the superior temporal sulcus (STS) of non-human primates.

call this condition delirium. Unless it is treated, patients can slip into a coma and die.

When nerves cells send messages, they give off small electrical currents, which can be amplified and measured with an EEC A physician who is concerned that a patient may be having epileptic seizure may order an EEC Electroencephalography can also be used to measure brain arousal. Normally, the EEG records waves of (p.104)

                   Attention

Figure 4–10. Diagram of a sagittal (midline) section through the brain. Under the cerebral hemispheres is an area called the midbrain or mesencephalon. The midbrain contains a large group of nerve cells called the reticular activating system. The reticular activating formation is located on both sides of the midbrain. The cells in this area are important in activating the higher portions of the brain including the cerebral cortex (broken lines with arrows). Stimulation of the reticular activating system arouses a sleepy animal, and injury to both sides of this system causes coma from which a person cannot be awakened. This diagram also demonstrates the position of a hemorrhage that injured the right side of the reticular formation. This lesion caused the right hemisphere to be unaroused and produced severe left-sided neglect.

electrical currents. In general, the more rapid the rate of wave action, the greater the arousal. Moruzzi and Magoun recorded EEGs in cats when they stimulated the reticular activating system (Fig. 3–3 and 4–10) and found that the cats not only became alert with stimulation, but their EEG waves also became more rapid. When Bob Watson and I reviewed this classic paper, we noted something interesting about which Moruzzi and Magoun did not comment. When they stimulated the cats’ reticular activating system on one side, the EEG recorded from the same side of the brain showed more arousal (the brain waves became more rapid) than did the EEG recorded from the opposite side of the brain. This suggested to us that if the reticular activating system on one side was injured, that side of the brain might become unaroused, or comatose, and unable to process stimuli from the opposite side of space.

To test this hypothesis, we made small lesions on one side of (p.105) monkeys’ reticular activating systems (Fig. 4–10). Postoperatively, these animals demonstrated the most severe neglect we have ever seen. We performed EEGs and found that the EEG rhythms recorded from the hemisphere with the reticular lesion were very slow.

Normally, we become aroused because there is an important stimulus in a particular area of space. Stimulus relevance is determined in part by the frontal-cingulate-parietal network we have discussed. Several investigators have stimulated different areas of the cerebral cortex to see which ones induce the greatest degree of arousal. They found that the cortical areas that influence arousal most intensely are the dorsolateral frontal lobes, the cingulate gyrus, and the parietal lobes. These are the same areas that we believe are critical in making attentional computations. Based on this set of observations, we proposed that a cortical-frontal-parietal-limbic (cingulate)-reticular network is important in mediating attention to stimuli on the opposite side of space (Fig. 4–11).

Right-Left Asymmetries of Attention

Many neurologists have noted that in patients with unilateral neglect, the lesion is much more likely to occur in the right hemisphere. Some have thought that it might only appear that neglect is more commonly associated with right-sided brain damage because patients with large lesions in the left hemisphere may be aphasic and unable to comprehend speech or reading, so that they cannot be adequately tested. Even when using nonverbal tasks without verbal

                   Attention

Figure 4–11. Diagram of the cortical (frontal and parietal)-limbic (cingulate gyrus)-reticular network, which is important in mediating attention.

(p.106) instructions, however, several investigators have found that neglect is both more common and more severe with right than with left hemisphere lesions. To account for this asymmetry, we suggested that perhaps the left hemisphere attends to stimuli primarily on the right side of the body and head, whereas the right hemisphere attends to stimuli in both left and right hemispace. Therefore, if the left hemisphere is injured, severe neglect does not occur because the right hemisphere can attend to ipsilateral right hemispace. However, when the right hemisphere is injured, the left hemisphere can attend to the right but not the left side of space, and the patient is inattentive to or unaware of stimuli in left space.

To test this hypothesis, Tom Van Den Abell and I studied normal college students with EEG, attaching electrodes to the area of the scalp over the right and left parietal lobes. The subjects sat at a table on which there was an apparatus with three lights. This apparatus was placed about 3 feet in front of the subject so that the middle light was directly in front of the subject's nose, the right light was in front of the subject's right ear, and the left light was in front of the subject's left ear. Directly in front of the subject's chest, we also placed a telegraph key. We told the subjects that their job was to press the telegraph key as soon as possible after they saw the middle light come on. We also told them that the lights on the right or left side might come on a little before the middle light and act as a warning that the middle light was going to come on.

The time period between the middle light coming on and the subject pushing the telegraph key is called the reaction time. Runners and swimmers know that if, before the starting gun goes off, a warning is given (“Take your mark, get set.…”), they will get off the starting block more rapidly than if no warning stimulus is provided. One reason the warning stimulus reduces reaction time is that it instructs the subject to attend to the starting stimulus. As predicted, our subjects’ reaction times were faster when the turning on of the middle light was preceded by either a right or left-sided warning stimulus than when there was no warning stimulus. The EEG taken while subjects were performing this task showed that when the warning stimulus was on the left side, the right hemisphere became activated. When the warning stimulus was on the right side, however, both hemispheres were activated. These results support the hypothesis that whereas the left hemisphere attends primarily to stimuli on the right side, the right hemisphere attends to stimuli on both sides. Therefore, if a person has a stroke in the left hemisphere, the right (p.107) hemisphere can still attend to the right side of space, but if the stroke affects the right hemisphere, the left hemisphere can attend only to the right side of space; the left side of space goes unattended.

Several years after our report of this experiment was published in Neurology, I went to a scientific meeting where a neuroscientist presented a functional imaging study. In this study, the researchers injected radioisotopes into subjects. The radioisotopes went to the active areas of the brain. The researchers used this technique to learn which areas of the brain are activated during an attentional task. They found that when the left side was stimulated, only the right hemisphere was activated, but when the right side was stimulated, both hemispheres were activated, suggesting that the left hemisphere attends to the right side of space and the right hemisphere attends to both the right and the left sides. The neuroscientist who presented this paper also mentioned that some investigators in a remote little southern town using EEG had described a similar observation.

When I first moved to this “remote little southern town” called Gainesville, it took me only 10 minutes to drive from my home to the hospital. As the town grew and traffic got worse, the trip began to take 20 or even 30 minutes. I am one of those people who constantly talks to themselves, especially when heavy traffic makes driving slow. On the way to work one morning, I was speculating about why the brain was designed so that the left hemisphere attends to the right side of space and the right hemisphere attends to both sides. Just then, a car on my right side drove through a stop sign without stopping. I immediately attended to this car, swerved, and just missed being hit. After calming down, I started ruminating again and then realized why the brain might have this hemispheric attentional asymmetry. Since the left hemisphere mediates speech-language, when a person is talking the left hemisphere cannot be vigilant for stimuli on the right, but because the right hemisphere can attend to both sides of space, it can attend to cars on the right while the left hemisphere is busy talking. Specialization permits parallel processing.

SPOTLIGHTS AND FLOODLIGHTS OF ATTENTION

Evelyn George was born and raised on a ranch just south of Saint Mary's River, which separates eastern Florida from eastern Georgia. She rarely saw doctors. Once when she was in a pharmacy in (p.108) Jacksonville, buying inexpensive reading glasses, store personnel were checking customers’ blood pressures. Her blood pressure was found to be 200/120. The young woman who took her blood pressure was so alarmed that she took it again and found that it was about the same. She told Mrs. George that her blood pressure was very high and that she needed to see a doctor as soon as possible. Driving in Jacksonville made Mrs. George nervous. She did not have a family doctor, had no insurance, and did not follow this advice. Several weeks after her blood pressure was taken in the pharmacy, she drove her old pickup truck down to a farm outside of Gainesville to visit her son and granddaughter. After dinner, she sat down on the couch to watch television and seemed to doze off. When her son had the coffee ready, he tried to wake her but had trouble doing so. When she finally opened her eyes, she seemed confused. He called 911.

When the paramedics arrived they took her blood pressure, which was now 230/150. They transported her to Shands Hospital at the University of Florida, where she was admitted and sent to the intensive care unit (ICU) to receive intravenous drugs to reduce her blood pressure. The doctors thought her high blood pressure was causing her confusion, a condition called hypertensive encephalopathy (encephalo brain; pathy, sick). However, when her blood pressure returned to the normal range, she still appeared confused. The doctors then thought that she might have suffered an intracerebral hemorrhage, one of the most serious complications of severe hypertension. They performed a CT brain scan, which showed no hemorrhage but did show two cerebral infarctions, one in the superior parietal lobe of each hemisphere (Fig. 4–12). Severe hypertension often causes the blood vessels in the brain to go into spasm, which prevents them from bringing sufficient blood to areas of the brain. The doctors in the ICU called for a neurology consultation, and the patient was transferred to the neurology service.

Bob Watson, the attending neurologist, noted that Mrs. George's main problem involved visual recognition and asked me to examine her in our weekly neurology grand rounds. When I tested her visual acuity it seemed excellent, as she was able to read quite small letters. To see if she had lost vision in a specific spatial sector, I had her fix her eyes on my nose while I wiggled my finger in different areas of space and asked her to point to my finger when it moved. She performed flawlessly. Then I asked her to touch my finger with her index finger. She had trouble with this simple task, (p.109)

                   Attention

Figure 4–12. Diagram of Mrs. George's CT scan showing two cerebral infarctions (strokes), one in the left superior parietal lobe (SPL) and the other in the right superior parietal lobe (FL = frontal lobe, OL = occipital lobe, IPL = inferior parietal lobe).

often missing my finger. This reaching disorder is called optic ataxia. It is caused by lesions in the parietal lobes that extend deep into subcortical structures. The cerebral cortex, or gray matter, contains the nerve cells and the short connections between them. Under the cortex (subcortical region) is the white matter of the brain. The subcortical region contains axons that resemble electrical wires coming from the bodies of nerve cells situated in the cerebral cortex. These axons connect one area of the brain with others. Geschwind thought that parietal lobe lesions caused optic ataxia because they prevented visual information, which comes from the occipital lobes situated in the posterior part of the brain, from reaching the motor neurons located in the posterior portion of the frontal lobes. Without this visual information, the motor cortex cannot correctly control arm movements.

There is, however, another explanation for optic ataxia. Research in monkeys and humans suggests that the region of the parietal cortex where Mrs. George's infarctions occurred is important for spatial analysis. The spatial analysis performed by the parietal lobes might be similar to analyses students perform on graph paper when they take a course in analytic geometry. To graph the position of an object in space in relation to your body, you may want to use a three-dimensional grid with three axes: horizontal (right to left), (p.110) vertical (up and down), and radial (front to back). The region directly in front of your eyes could be the area where all three axes meet. This is called the zero point. An object to the right of your midline might be given a positive value, and an object to the left a negative value. Similarly, if an object is above or in front of your zero point it would have a positive value, and if below or behind it, it would have a negative value. In this graph, if each unit equaled 1 inch and an object's position had coordinates of -30, + 40, and +20, this would mean that the object is 30 inches to the left of midline, 40 inches above eye level, and 20 inches in front of your eyes. Some investigators believe that when you see an object in space, it is the parietal lobes that compute these coordinates and then supply this information to other regions of the brain, such as the motor areas that are important in programming movements of the arms. If this part of the brain is injured, as it was in Mrs. George, the individual cannot compute the locations of objects in space and thus has optic ataxia.

Mrs. George also had difficulty moving her eyes (ocular ataxia or psychic paralysis of gaze) to fixate on the target at which she wanted to gaze. Like the hands and arms, the eyes are directed by a motor system. The motor system that controls eye movements is independent of the motor system that controls arm movements, but this oculo-motor system also relies on the computations of the spatial coordinate system that was injured in this woman.

Although Mrs. George had optic ataxia, once her eyes found the object she was able to continue to gaze at it. She could name objects I showed her, but when I presented a picture of a family eating dinner and asked her to describe it, she said it was a picture of an apple. Since her visual acuity was excellent and she could see my hands in all portions of space (normal visual fields), her response surprised me. I thought that perhaps she did not understand me, so I asked her again to tell me what the picture was illustrating. She turned, looked at me, looked back at the picture, and again said, “This is a picture of an apple. See, here is the apple.” She then pointed to an apple on the table in the picture but ignored the rest of the picture. Someone who was attending grand rounds found another complex picture. This was a painting of a Civil War battle. It looked like the Battle of Olustee, which took place very close to Mrs. George's home. She looked at the picture and said, “This is a picture of a bird.” She then pointed to a bird that the artist had placed (p.111) above the battle scene. “See here, there is the bird that is flying.” She never mentioned anything else.

This disorder, in which the person sees only a small part of a complex picture, is called simultanagnosia (simultan = simultaneous; a = without; gnosia = knowledge). The combination of simultanagnosia, optic ataxia, and ocular ataxia is called Balint's syndrome, named after the neurologist who first described it in 1909. The mechanism that accounts for the inability to see the whole picture is not known. For some tasks, we need a narrow focus of attention, or spotlight; for others, we need a broad view, or floodlight. See Figure 4–13 for an example of a task that may require a floodlight or a spotlight. Although a floodlight and a spotlight may have equal wattage, the lens on the spotlight focuses light in a small area and the lens on the floodlight distributes light to a large area. Anna Barrett and other investigators in our laboratory have found that the parietal lobe is important in controlling how narrowly or widely attention should be focused for a specific task. Perhaps Mrs. George couldn’t use her attentional floodlight, depended on her attentional spotlight, and missed the “big picture.”

The Navon figure in Figure 4–13 is designed to pit these two attentional lenses against each other. If you look at this figure and first see the letter H, this might suggest that you tend to use your attentional floodlight first. In contrast, if you see the letter A first, this might suggest that you tend to use your attentional spotlight. After giving patients with unilateral brain damage a similar task, Lynn Robertson and her coworkers found that whereas the left

                   Attention

Figure 4–13. Navon figure. A Patient who has a problem with the attentional floodlight may be unable to find the large letter H. A patient who has a problem with the attentional spotlight may be unable to find the letter A.

(p.112) hemisphere is important in mediating the spotlight, the right hemisphere is important in mediating the floodlight.

HABITUATION

When you were reminded about your left foot you may have attended to it and felt it for a few minutes, but before you started reading this section you may have again become unaware of that foot. You became unaware of it because there were no novel stimuli impinging on it and nothing was happening to it that needed your attention. The phenomenon of becoming unaware of irrelevant stimuli is called habituation. You can habituate in all sensory modalities. For example, when you walk into a room to read, you may hear the air conditioner and smell an odor, but a few minutes after you start to read, you may no longer be aware of them.

While testing Mrs. George's vision I observed another defect that I’d never seen before. I held up a pen and asked her what it was. She responded, “It's a pen.” The I asked her to tell me as much about the pen as she could (e.g., What kind of pen is it? What are its colors? Does it have a clip?). She then said, “Dr. Heilman, I know this sounds funny, but I cannot see the pen. It disappeared.” I thought something had happened to her vision. Perhaps another stroke had destroyed her primary visual areas, so I put the pen down because I wanted to examine her vision further. When I moved the pen, she said, “There, now that you moved the pen, I can see it again.” I didn’t know if her visual cortex was temporarily deprived of blood or if this phenomenon was a result of her previous stroke. To continue the test, I held up my keys and asked her to stare at them. After a few seconds she said, “The keys disappeared.” When I moved the keys, she said, “There they are again.” I asked how long she had been having this problem with disappearing objects. She said, “Since I had my stroke, but I didn’t tell because I am worried that the doctors would think that I’m crazy.”

Mark Mennemeier, a postdoctoral fellow in our laboratory at that time, told me that this phenomenon of objects disappearing can even be seen in normal people. For example, if you look at the cross on the right side of Figure 4–14, at first you will see the dot on the left side of the page. However, if you keep staring at the cross without moving you eyes or the paper, after about 20 to 30 seconds the dot will either fade or disappear. This fading phenomenon in (p.113)

                   Attention

Figure 4–14. If you stare at the X on the right side of the page without moving your eyes or the page, after about 30 to 60 seconds the dot on the left side of the page may either disappear or fade. This phenomenon is called sensory habituation. After the dot fades, if you move the page it will reappear.

normal people is called the Troxler effect. If you move either your eyes or the paper, the dot will reappear. What was abnormal about Mrs. George was that the disappearing items were in her central vision, and they disappeared very rapidly.

In the center or core of the brain is a sensory relay structure called the thalamus (Fig. 4–15). Sensory information coming from the visual, auditory, and tactile systems must pass through the thalamus before entering the cerebral cortex for analysis. Each of the sensory systems courses through a different area of the thalamus.

                   Attention

Figure 4–15. Diagram of the thalamus, which is located below the cortex in the center of each hemisphere. On the onside, surrounding the thalamus, is a group of nerves called the thalamic reticular nucleus. These cells normally inhibit visual (or other sensory) stimuli from traveling to the cerebral cortex for further processing. The parietal lobes appear to inhibit (-) these inhibitory cells and the frontal lobes activate (+) these inhibitory thalamic reticular neurons, thereby blocking the transmission of sensory information to the cerebral cortex. In this figure the inhibitory cells (-) are dark and the excitatory cells (+) are light.

(p.114) Surrounding these areas of the sensory thalamus is a neural structure called the thalamic reticular nucleus. This structure has nerve cells that project to the sensory relay areas of the thalamus. The thalamic reticular nerves are inhibitory and, when activated, they prevent the thalamic sensory areas from transmitting sensory information to the cortex. Therefore, the thalamus is not only a relay station but also may be a gate that controls which sensory information reaches to the cerebral cortex. Based on our observations of Mrs. George, we proposed that areas of the cerebral cortex may control this sensory gate, either opening or closing it. That objects disappeared rapidly for Mrs. George suggests that her sensory gate was closing prematurely. Since she had parietal lobe lesions, we proposed that normally the parietal lobes may open the gate by inhibiting this inhibitory thalamic reticular nucleus. We also thought that the frontal lobes might be important for closing this gate. To test this hypothesis, Mark Mennemeier and others in our laboratory used the Trox-ler test with patients who had frontal and parietal lesions. Our findings support this hypothesis. One patient, with a parietal lesion in one hemisphere and a frontal lesion in the other hemisphere, was particularly interesting. Using a Troxler fading test, we found that when the dot was on the side of space opposite the parietal lobe lesion, the patient experienced fading more rapidly than did control subjects. When the dot was on the side of space opposite the frontal lobe lesion, fading took much longer than it did in the controls or did not occur. These observations indicate that whereas the parietal lobes are important in helping us attend to novel or important stimuli, the frontal lobes are important in helping us withdraw attention from insignificant stimuli.

SUMMARY

The brain has a limited processing capacity and receives more information than it can process fully. We attend to stimuli that are important and ignore those that are unimportant. Stimulus significance is determined by our immediate needs (drives) and future goals. When the inferior parietal lobe, the dorsolateral frontal lobe, or the cingulate gyrus is injured, patients are inattentive to stimuli on the opposite side of space. In addition to receiving projections from the sensory association cortex (visual, auditory, and tactile), (p.115) the inferior parietal lobe has reciprocal projections with the dorsolateral frontal lobe and the cingulate gyrus. The frontal lobes are important in mediating goal-oriented behavior and the cingulate gyrus, a portion of the limbic system, is important in mediating drives and emotional behaviors. Thus, a frontal lobe-cingulate gy-rus-inferior parietal lobe network is critical in determining stimulus significance. When a person attends to an important stimulus, he or she becomes more alert or aroused. Arousal increases the brain's processing capacity. The midbrain (mesencephalon) contains a group of neurons called the reticular activating system, which when stimulated produces cortical activation or arousal. The dorsolateral frontal lobe, the cingulate gyrus, and the parietal lobe control this activating system. The cortical-(frontal-parietal)-limbic (cingulate)- reticular network, important in mediating attention-arousal, is summarized in Figure 4–11. Although each hemisphere contains one of these networks, die network in the right hemisphere appears to be dominant. In addition, some tasks require us to focus attention on a small area of space (attentional spotlight), and other tasks require us to attend to a large area of space (attentional floodlight). Studies of patients with posterior temporal-inferior parietal lesions suggest that the left hemisphere may be more important for focused attention and the right for distributed attention. Finally we become unaware of irrelevant stimuli (habituation). The frontal lobes might be important in closing the gate in the thalamus, thereby preventing stimuli from reach in cerebral cortex, and the parietal lobe might be important in opening this gate.

Selected Readings

De Renzi, E (1982) Disorders of space exploration and cognition, Wiley, New York.

Heilman, K.M., Watson, R.T., Valenstein, E. (1993) Neglect and related disorders. In Clinical neuropsychology, Heilman, K.M. and Valenstein E. University Press, Oxford, New York, pp. 279-336.

Heilman, K.M., Valenstein, E, Watson, R.T. (1983) Localization of neglect. In Localization in neurology. (Ed.) Kertesz A. Academic Press, New York, pp. 471-492.

Jeannerod, M. (1987) Neurophysiological and neuropsychological aspects of spatial neglect. Elsevier, Amsterdam.

Pardo, J.V., Fox, P.T., and Raichle, M.E. (1991) Localization of a human system for sustained attention by positron emission tomography. Nature 349:61–64.

(p.116) Posner, M.I., and Rafal, R.D. (1987). Cognitive theories of attention and rehabilitation of attentional deficits. In Neuropsychological rehabilitation, (Ed.) Mier, M.J., and Benton, A.L., and Diller, L. Guilford Press, New York, pp. 182–201

Roberson, I.H., Marshall, J C, (1993) Unilateral neglect: Clinical and experimental studies. Lawerence Erlbaum, Hove, United Kingdom.