| ||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 12
Opportunities in Neuroscience for Future Army Applications
2
Neuroscience and the Army
This chapter provides a brief overview of neuroscience and neuroscience technology, including the definition of neuroscience used by the committee for the study. It describes major applications areas that were considered by the study and provides rationale for not considering some applications.
HISTORY, SCOPE, AND DEFINITION OF NEUROSCIENCE
“Neuroscience” refers to the multiple disciplines that carry out scientific research on the nervous system to understand the biological basis for behavior. Modern studies of the nervous system have been ongoing since the middle of the nineteenth century. Neuroanatomists studied the brain’s shape, its cellular structure, and its circuitry; neurochemists studied the brain’s chemical composition, including its lipids and proteins; neurophysiologists studied the brain’s bioelectric properties; and psychologists and neuropsychologists investigated the organization and neural substrates of behavior and cognition.
The term “neuroscience” was introduced in the mid-1960s, signaling the start of an era when these disciplines would work together cooperatively, sharing a common language, common concepts, and a common goal: to understand the structure and function of the normal and abnormal brain. Neuroscience today spans a wide range of research endeavors, from the molecular biology of nerve cells, which contain the genes that command production of the proteins needed for nervous system function, to the biological bases of normal and disordered behavior, emotion, and cognition, including the mental properties of individuals as they interact with each other and with their environments.
Neuroscience is one of the fastest growing areas of science, and the brain is sometimes referred to as the last frontier of biology. In 1971, for example, the first meeting of the Society for Neuroscience was attended by only 1,100 scientists; in 2007, 26,000 scientists participated at the society’s 37th annual meeting and more than 15,000 research presentations were made. At this time, national societies of neuroscience exist throughout the world, and there exists a Federation of European Neuroscience Societies.
Neuroscience incorporates a number of interacting areas, including cognitive neuroscience, systems neuroscience, cellular and molecular neuroscience, developmental neuroscience, clinical neuroscience, theoretical neuroscience, and computational neuroscience. Operations involving neurons1 form the basis for all of these areas and take place on four fundamental hierarchical levels: molecular, cellular, systems, and behavioral. These levels rest on the principle that neurons communicate chemically by the activity-dependent secretion of “neurotransmitters” at specialized points of contact called “synapses.” In order for the brain to perform its multiple functions, the mental operations of the brain rely on a properly functioning and integrated system of autonomous bodily functions, monitored by the brain and modified by behavioral operations only when the autonomous regulatory systems—thermal control, control of blood nutrients, orientation of the body and limbs in space while moving, gesturing, and transporting, control of the salt and water balance, and so on—are compromised.
The autonomous peripheral systems monitored by the brain and over which it has ultimate control allow for an extreme but generally subconscious interaction between the body’s physical fitness and the brain’s emotional and cognitive powers to drive the body’s performance under demanding conditions. For example, overcoming fatigue through personal willpower alone is a learnable ability that is characteristic of superior athletes and that would be beneficial for military personnel as well (see Chapter 5).
Many fields of clinical medicine are directly concerned with the diseases of the brain. The branches of medicine most closely associated with neuroscience from the perspectives of this report are neurology (the study of the degenerative
1
Nerve cells of the central and peripheral nervous systems.
OCR for page 13
Opportunities in Neuroscience for Future Army Applications
sensory and motor diseases of the brain), neurosurgery (the study of the surgical treatment of neurological disease), and psychiatry (the study of behavioral, emotional, and mental diseases). Other fields of medicine also make important contributions to neuroscience, including neuroradiology, which is the use of radiation for imaging the brain—initially with X-rays and, more recently, with positron emitters, radiofrequency, and electromagnetic waves—for clinical studies and microscopic study of samples from diseased neural tissue.
Hierarchical Levels of Neuroscience
At the molecular level, one examines the interaction of molecules—typically proteins—that regulate gene expression and translation into proteins. Proteins mediate neurotransmitter synthesis and storage and release other essential neuronal molecular functions such as the receptors by which neurons respond to neurotransmitters. Most drugs used for the treatment of neurological or psychiatric diseases work by either enhancing or diminishing the effects of neurotransmitters.
At the cellular level of neuroscience, one examines the interactions between neurons through their synaptic connections and between neurons and the supporting cells, the glia. Research at the cellular level strives to determine the neural pathways by which specific neurons are connected and which of their most proximate synaptic connections might mediate a behavior or behavioral effects of a given experimental perturbation.
At the systems level, one examines the interconnected neural pathways that integrate the body’s response to environmental challenges. The sensory systems include the specialized senses for hearing, seeing, feeling, tasting, and balancing the body. The motor systems control trunk, limb, eye, and fine finger motions. Internal regulatory systems are responsible for, among other things, control of body temperature, cardiovascular function, appetite, and salt and water balance.
At the behavioral level of neuroscience research, one examines the interactions between individuals and their collective environment. Research at this level centers on the systems that integrate physiological expressions of learned, reflexive, or spontaneous behavioral responses. Behavioral research also looks at the cognitive operations of higher mental activity, such as memory, learning, speech, abstract reasoning, and consciousness. Research over the past three decades has established that the brain is highly adaptable (this ability is commonly termed “neuroplasticity”) at each level of operation: the activity-dependent ability to change gene expression, to change transmitter production and response, to change cellular structure and strength of connections between neurons, and to change behaviors by learning.
An important consequence of organizing neuroscience research at four vertical hierarchical levels is that it enables one to hypothesize experimental results on one level based on experimental findings and observations from other levels. This ability extends to hypothesizing neuronal operations or neuronal diseases based on data that would predict results at the behavioral level given the results of a perturbation or other experimental manipulation at a lower level. Such results are strongly supported in the literature (Aston-Jones and Cohen, 2005a, 2005b).
One might predict from experimental results in animals, for example, that the thin axons that establish functional properties of the noradrenergic system might also be one of the brain fiber systems most vulnerable to the percussive damage of traumatic brain injuries (TBI), such as might result from an improvised explosive device (IED). As discussed in the Chapter 5 section on brain injury, this is indeed the case, and the ability to translate between levels of neuroscience has proven helpful in the treatment of TBI and its emotional effects.
NEUROSCIENCE TECHNOLOGIES
Until the advent of modern computer-based technology, the primary noninvasive tools used to understand the workings of the central and peripheral nervous system were the recording of electrical signals from the scalp (electroencephalography [EEG]) and X-ray imaging of the soft tissue of brain as distinguished from bone and compartments containing cerebral spinal fluid (CSF). EEG allowed detecting epileptogenic foci that could subsequently be managed surgically if a discrete region was involved in the initiation of seizures or pharmacologically if the region was more generalized. The X-ray imaging allowed detection and localization of lesions because the lesions displaced readily identified portions of the brain. However, these technologies provided very limited insight into neural information processing related to cognition, the central mechanisms involved in the perception of pain, or other higher-order brain activities. The pioneering work of Penfield and his colleagues was an exception: It combined EEG with invasive brain surgery to associate the visual and auditory auras that accompanied seizures to specific regions in the visual, auditory, or temporal cortices (Penfield and Perot, 1963).
The two decades from the late 1980s to the present have seen the rapid rise of technologies that can provide high-resolution structural images of the gray and white matters of brain as distinct from one another, clearly delineating details as small as the foci of white-matter disease and inflammatory changes. These technologies are capable of imaging the metabolic processes that are associated with functional activity of the brain in response to specific stimuli (positron emission tomography [PET] and functional magnetic resonance imaging [fMRI]); the orientation and dimensions of axonal fiber bundles connecting one brain region to another (diffusion tensor imaging); and the electrophysiological localization of brain activation (magnetoencephalography
OCR for page 14
Opportunities in Neuroscience for Future Army Applications
FIGURE 2-1 Various noninvasive imaging technologies provide insight into the brain (anatomy) and mind (function). Resolution boundaries are approximate. Currently, high resolution requires invasive procedures or injection of pharmacological agents not approved by the Food and Drug Administration; such technologies are outside the scope of Army applications, but some discussion is provided in Chapter 7. Note that useful measurements are performed at any point in the brain-mind plane. SOURCE: Adapted from Genik et al., 2005.
[MEG] and visual and auditory evoked potentials). Most of the newer methodologies depend on computer-driven signal averaging across a time span of seconds to provide an understanding of brain structure and function.
The advancement of noninvasive neural measurement techniques has opened new windows to the study of the functioning human brain. Figure 2-1 illustrates how neuroscience technologies provide insight into the brain (anatomy) and the mind (function). The spatial resolution of a given technology defines the largest and smallest brain structures that can be observed, while the temporal resolution defines the elements of mind function to be measured. Academic and commercial research are primarily geared to improving resolution, although important measurements for the prediction of behavior can be made at any point in the brain-mind plane. Chapter 7 contains further details of neuroscience technologies, including a discussion of more invasive modalities and the future direction of noninvasive imaging research.
The spatial resolution of the structural imaging modalities that could be achieved with the human brain in 2008 approximates 100 microns per side in three-dimensional pixel elements when a 7-tesla (T) MRI system is employed, which is the highest-field whole-body scanner currently in commercial production. This resolution requires a total imaging time of less than 10 minutes in a series of planes. Recognizing that a large neuron cell body has a diameter of 20-30 microns, each 100-micron element is likely to contain eight neurons and a much larger number of glia. Also recognizing that each dendrite of a large neuron (Purkinje cell or motor neuron or pyramidal cell) may contain 103 to 104 synaptic connections, the number of information-processing elements contained within the small pixel volume is large. There are very few 7-T MRI instruments available for human imaging at this time. The majority of MRI machines are 1.5 T and 3 T. The resolution for these instruments nominally decreases proportionally with static field, assuming a similar image acquisition time. Longer imaging times can yield higher resolution, but the trade-off is that any movement of the head by the subject markedly decreases the resolution. The limits of structural resolution are therefore spatiotemporal limitations.
The diffusion tensor imaging allows examining the pathways along which groups of axonal fibers travel from one brain region to another and the vectorial direction of the fiber bundle group. This information is essentially a mapping of the potential for information exchange between regions and the tracing of the bundles that could carry information. No information is provided about the actual informational
OCR for page 15
Opportunities in Neuroscience for Future Army Applications
transactions between brain regions or the temporal events associated with such transactions.
The MRI system depends on alignment of the bulk water dipole in the magnetic field and the perturbation of this alignment by a radiofrequency pulse. The realignment of the dipole following the radiofrequency pulse provides the signal of interest. What MRI then permits resolving is the relative water content of each compartment in the brain, spinal cord, or organ system that is being studied. The white matter has the least free water content, the gray matter has intermediate levels, and the CSF compartments have the greatest amount of water. These compartments can be readily resolved. The loss of myelin accompanying multiple sclerosis or other white-matter disease is also readily discerned for this reason, as are other diseases due to neurodegenerative changes and cell loss. (MRI is to soft tissue as a computerized axial tomography [CAT] scan is to bone. These methods are essentially noninvasive.)
Imaging of the brain based on X-ray methods (CAT scans or X-ray scans) also have a high resolution but do not readily differentiate white from gray matter as does an MRI device. This approach has proven to be very useful for threading catheters intravascularly to within 1 mm of any location in the brain, for localizing tumors in the cranial cavity and region of the spinal cord, and for visualizing new blood from hemorrhages. The two modalities taken together offer exquisite timing of tumor, infection, and bleeding developments.
The technologies for structural imaging provide the basis for understanding the brain regions of interest obtained by functional images. Functional images are lower resolution than structural images in order to facilitate acquiring many repetitions in a short time. Typical 3-T fMRI acquires 2- to 3-mm isotropic resolution whole-brain images every 1 or 2 seconds, while cutting-edge hardware can acquire 1-mm isotropic resolution whole-brain images in 1 second. By acquiring the images in a time-locked fashion with stimulation, one can interpret the activation of brain regions as a response to stimulation. The fMRI signal is generated by the loss of oxygen from hemoglobin (the deoxygenated hemoglobin is paramagnetic, while the oxygenated is not) and the change in blood flow in the activated region. The blood flow response continues for several seconds, enabling detection of very fast physiological events using this rather slow temporal sampling. This is called blood oxygen level-dependent (BOLD) imaging.
The more repeatable the brain response and, thus, the blood flow response, the better localized activations can be observed, because one can average over several trials. If one can constantly stimulate a response in a block of time, say the visual cortex with a flashing checkerboard, an activated state lasts longer and is easier to detect. However, some cognitive processes are inherently transient and are better observed in so-called event-related experiments. Which experimental method to use depends on the physiology of the reaction one wishes to study (Friston et al., 1999; Otten et al., 2002).
In PET imaging, a solution containing a small amount of a radioactive element is injected into the subject’s bloodstream. Signals from the decaying isotopes will show localized concentration, which could be due to increased blood flow, say, or perhaps to glucose metabolism, depending on the tracer used. PET functional images are acquired at 3- to 4-mm isotropic resolution, and because very little tracer is used (to protect the subject), they require a few minutes to tens of minutes of averaging to show a significant signal, again, depending on the tracer used. The significant advantage of PET over fMRI is that specific metabolic reactions can be targeted for measurement rather than merely observing changes in blood flow. Additionally, with proper calibration, an absolute measure of change in brain function can be determined, even in different scanning sessions, whereas with fMRI only changes relative to a baseline condition are observable. The main disadvantage of PET is that the physiology that produces transient signals of short duration, akin to noise, cannot be observed. Moreover, to avoid any adverse effects of radiation on the subject, PET studies on a given individual are limited to three or four experiments per year depending on the amount of emitter injected. The functional properties of the brain have been shown to have relevance for detecting a loss of situational awareness in individuals subjected to total sleep deprivation for more than 24 hours (Belenky et al., 2003).
Some of the noninvasive imaging technologies have characteristics that may limit their use in military applications. PET, for example, requires injection of a positron emitter into the human subject. On the other hand, fMRI is truly noninvasive in that no contrast agents or radioisotopes are injected in the subject.
A recent study by Bakker et al. (2008) demonstrated that activation of the fMRI in hippocampal regions was related to the detection by human subjects of minor changes to objects in the visual field. Detection of changes in the visual field by an unconscious mind might be applied to the identification of IEDs implanted in the terrain. A clear understanding of the mechanisms by which an individual can become alerted to such activations when hippocampal neurons fire might enhance the mission success of Army personnel.
Electrophysiological events accompanying brain stimulation can be recorded in real time (milliseconds) from the cortex but not from deep brain structures, and the spatial resolution is 5 mm per recording element. For noninvasive studies with the calvarium intact, the recordings are essentially averaging events in the top millimeter of cortex and across a diameter of 5 mm or more. The evoked potential is a signal-averaging methodology that examines rates of event-related depolarization across spatial domains of the scalp associated with a specific stimulus presentation—for example, a checkerboard flicker, a flash, or an auditory signal. It is possible to detect changes in such events that are due to changes in information processing (or to central nervous system lesions). Because vectorial electrical events are
OCR for page 16
Opportunities in Neuroscience for Future Army Applications
BOX 2-1
Computational Processes in the Human Brain
The last half century has seen the emergence of multiple technologies that, when taken in total, can lead to new understanding of the human–machine interface and to improving our management of it as an interacting organism. The relatively concurrent development of high-speed computer data processing, anatomic mapping of the human brain, noninvasive imaging technology (capable of visualizing neural information processing at a resolution of 1 mm3 in three dimensions), decoding of the human genome, and nanotechnology (giving us the ability to prepare molecular entities that can self-assemble in predictable arrays) has permitted researchers and theoreticians to construct models of how human–machine interfaces can aid training, increase combat effectiveness, and speed up the acquisition of information by the mind. A summary of some of this can be found in Kurzweil (2005, pp. 122-128).
Anatomic studies of the brain lead to an estimate that the brain contains about 1011 neurons (information-processing units) and that many neurons (Purkinje cells in the cerebellum or anterior motor neurons in the spinal cord) have 1,000 information input elements (synapses) and 10 output elements (synapses), yielding about 1014 possible information transactions occurring in seconds (it is assumed that in a local circuit in the cerebellum there are rapid impulse and repolarization events). If we focus on the neural components that enable all human visual perception, the retina is the initial image capture and processing element. It is a structure about 2 cm wide by 0.5 mm thick and weighs about 20 mg. This volume of tissue will contain about 105 neurons. Local information processing occurs between the rod and cone cells, which are activated by the input from the visual field, and the ganglion cells, which transmit initially processed information to the occipital cortex of the brain for further processing.
The retina provides initial processing capability such as rapid detection of edges of objects by center-surround inhibition/activation, the movement of elements in the visual field, and the perception of dark/light properties of the objects in the field. The range of light/dark information processing that can be detected is several orders of magnitude. During the next two decades it is likely that visual input devices will be developed that will be able to fuse visual data with auditory cues to allow the more rapid detection of threats. These devices will be able to sustain vigilance by fusing measures of decreases in situational awareness (obtained through electroencephalography (EEG), evoked potentials, or other cues from the soldier) with threats detected in the area of operations.
accompanied by changes in the magnetic field perpendicular to the electrical path, it is possible to measure columnar electrical events in the sulci using MEG.2 The spatial resolution with this method is on the order of centimeters, and the temporal resolution is on the order of fractions of seconds to seconds. Studies on the effects of various stressors on cognitive ability using MEG with EEG provide evidence that this tool can detect decremented situational awareness. This is important from a defense perspective since EEG-related devices are portable and readily deployable on an individual soldier while MRI devices are not (Box 2-1).
Among the applications of modern neuronal imaging and structural technologies to military needs are the assessment of (1) when appropriate training-to-criterion on a given skill set has been achieved; (2) when a soldier in the well-rested state has significantly degraded situational awareness of her/his capability; (3) early signatures of neural dysfunction. While some of these methodologies may prove to have application for the a priori assessment of highly desirable traits such as leadership, persistence, and other successful warrior behaviors, there is little evidence to support application of the neuroscience tools currently available to tasks such as the three mentioned above.
The report thinks of technology as being in one (or sometimes both) of two categories: technologies that are “mission-enabling” (deployable) instruments and those that are “research-enabling” instruments. The word “instrument” is used in the most general sense: It could refer to a pen-and-paper personality inventory, a software-controlled skills survey, a reaction-time analysis method for training assessment, a control interlock system to distribute information to vehicle crew based on their current workload and baseline cognitive capability, an in-helmet device designed to monitor neural activity or cerebral blood flow, or a device that advances an imaging technology. Neuroscience research plays a major role in the development of instruments in both categories of technology.
RELIABLE BIOMARKERS FOR NEUROPSYCHOLOGICAL STATES AND BEHAVIORAL OUTCOMES
Recordings of the brain’s electrical activity and the images of neural activation regions that were described in the previous section on neural monitoring technologies are all methods to record changes in brain activity in a specific location at a given time. Generally, these data on neural events are interpreted in terms of what they tell us not just
2
This is possible because in the cerebral cortex, information-processing modules are composed of a column of neurons spanning the thickness of the cortex.
OCR for page 17
Opportunities in Neuroscience for Future Army Applications
about that event but also about longer-lasting patterns of brain activity and brain structure—that is, about neural states and changes in neural state. A major, if not the primary, goal of studying neurophysiology with these techniques is to understand the linkages among neural states, psychological states, and behavior. (Psychological states are traditionally self-reported; behavior is directly observable by others.) In short, the outputs of neural monitoring technologies are indicators of neural state (and of changes in neural state) that can in turn be linked with behavior.
Typically, these linkages begin with relationships that have statistical significance on a group-averaged basis and move, as the state of scientific knowledge progresses, to connections that hold on an individual basis.3 Ultimately, this refinement will lead to a reliable, scientifically defensible, knowledge of necessary and sufficient causal conditions underlying and explaining the observed patterns of brain activity, mental experience, and behavior. But because that ultimate goal is still a long way off, care must be taken not to leap prematurely from a statistically significant correlation to conclusions about causality.
In addition to brain activity signals and images, the neurosciences and allied fields—ranging from genetics to molecular biology and traditional behavioral science—are exploring a wide range of phenomena that can be connected with neuropsychological states or changes in state. Just as a neuroimaging pattern can be used as an indicator or marker of neuropsychological state, so may other phenomena. Among the phenomena being studied for this purpose are biologically active small molecules, proteins and related molecules (e.g., lipoproteins and metabolic residues or precursors of proteins), genes and nonencoding regions of the genome, physiological events or patterns outside the brain but within the organism, and responses to an environmental exposure (physical, chemical, biological, social, or psychological). The variety and complexity of the hypotheses being put forward and tested about such correlations are driving a revolution in scientific understanding.
“Biomarker” is a term often used in the biomedical disciplines for a characteristic that can be used as an indicator of some biological condition or outcome that is ultimately of interest but difficult to ascertain directly, at least under conditions of interest to a particular application. A number of implicit and explicit definitions of “biomarker” are in common circulation.4 There are also quite different uses of the term in other disciplines.5 To avoid confusion, the committee has adopted the following definition, published by the Biomarkers Definitions Working Group of the National Institutes of Health (Atkinson et al., 2001, p. 91):
Biological marker (biomarker): A characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention.
In its report, the Biomarkers Definitions Working Group focused on applications of biomarkers as surrogates for clinical end points in a study or clinical trial. This emphasis on a biomarker’s role as a surrogate for a physiological or behavioral condition or outcome is evident in the Working Group’s definition of clinical end point as “a characteristic or variable that reflects how a patient feels, functions, or survives.” Common examples of biomarkers mentioned by the Working Group include elevated blood glucose concentration for the diagnosis of diabetes mellitus, the concentration of prostate-specific antigen in blood as an indicator of the extent of prostate tumor growth and metastasis, and blood cholesterol concentration as a predictive and monitoring biomarker for heart disease risk (Atkinson et al., 2001, p. 91).
Throughout the remainder of this report, the Committee on Opportunities in Neuroscience for Future Army Applications is primarily interested in biomarkers as objectively measured and evaluated indicators of either a neural state or a behavioral outcome. For example, Chapters 3 and 5 discuss the use of neuroimaging as a source of biomarkers for individual response to particular environmental stressors. Chapter 7 discusses the value for Army applications of finding biomarkers that can be measured under field conditions and that are reliable indicators of specific neural states that have been reliably linked to behavioral outcomes. These biomarker applications differ from uses of biomarkers as surrogate clinical end points, which were the focus of the Biomarkers Definitions Working Group. Nevertheless, the Working Group’s caveats about biomarker applications are useful cautions for any application.
For example, both the accuracy and precision of a biomarker as a surrogate measure of outcome must be demonstrated:
3
For example, fMRI investigations typically begin by examining relationships averaged over many trials per subject and then averaged over multiple subjects. Once a statistically significant relationship is established in this way, the typical next step is to show that the relationship holds for individual events in the group of subjects, and ultimately to individual events in each subject.
4
The National Cancer Institute defines a biomarker as “a biological molecule found in blood, other body fluids, or tissues that is a sign of a normal or abnormal process, or of a condition or disease …” in its Dictionary of Cancer Terms, available at http://www.cancer.gov/dictionary/. Accessed November 23, 2008. The MedlinePlus online medical dictionary uses a definition from Merriam Webster, in which a biomarker is “a distinctive biological or biologically derived indicator (as a biochemical metabolite in the body) of a process, event, or condition (as aging, disease, or exposure to a toxic substance).” The usage example given is “age-related biomarkers of disease and degenerative change.” The URL for this National Institutes of Health (NIH)-sponsored dictionary is www.nlm.nih.gov/medlineplus/mplusdictionary.html. Accessed on November 23, 2008.
5
In petroleum exploration, biomarkers are compounds found in geologic extracts (including oil, rock, sediment, and soil extracts) that indicate a biological origin of some or all of the material. See “Using Oil Biomarkers in Petroleum Exploration,” by Oiltracers LLC, available at www.oiltracers.com/biomarker.html8. Accessed November 21, 2008. A similar use occurs in planetary science and astrobiology, where a biomarker is a chemical that signals the presence of biological processes, often in the distant past.
OCR for page 18
Opportunities in Neuroscience for Future Army Applications
The utility of a biomarker as a surrogate endpoint [or as a surrogate for a neural state or behavioral outcome] requires demonstration of its accuracy (the correlation of the measure with the clinical endpoint [or the neural state or behavioral outcome]) and precision (the reproducibility of the measure). (Atkinson et al., 2001, p. 92; bracketed text added by the committee)
Elsewhere in its report, the Biomarkers Definitions Working Group notes as follows:
Biomarkers that represent highly sensitive and specific indicators of disease pathways have been used as substitutes for outcomes in clinical trials when evidence indicates that they predict clinical risk or benefit. (Atkinson et al., 2001, p. 90; emphasis added by the committee)
The attributes of sensitivity and specificity have rigorous definitions that can be applied across the range of physical, biological, and even social characteristics that are or will be candidates for indicators of whether a condition such as a neural state is present or not. In a binary test (in this case, whether the neural state of interest is or is not present), “sensitivity” is defined mathematically as the number of test instances in which the biomarker is positive and the neural state is present, divided by the number of instances in which the neural state was present whether or not the biomarker was positive. In other words, the measure of sensitivity is the ratio of true positive tests to the sum of the true positive and false negative tests. (False negative tests are those that should have been positive.) “Specificity” is defined as the number of instances in which the biomarker was negative and the neural state was absent, divided by the number of instances in which the neural state was absent whether or not the biomarker was present. The measure of sensitivity is thus the ratio of true negative tests to the sum of the true negative and the false positive tests. A reliable biomarker for this kind of binary application is one that has both high sensitivity and high specificity; that is, both ratios are close to unity.
With respect to how biomarkers may be used, they can be current, retrospective, or predictive (prospective) indicators or measures, depending on whether the condition or end point for which they are used as a surrogate occurs at the same time, before, or after the assessment of the biomarker. Demonstrating the reliability of a biomarker for an application typically requires the same temporal relationship as the intended application.
As the report by the Biomarkers Definitions Working Group also notes, often several biomarkers must be combined to get a reliable indicator or measure of outcome (Atkinson et al., 2001, p. 93). For applications of practical value to the Army, such as field-deployable indicators of neural state, this approach, assessing multiple biomarkers, may often be required (see Chapter 7 for further discussion).
In summary, applications of biomarkers as surrogates for neural states or behavioral outcomes require a demonstration of reliability that exceeds mere statistical correlation. In cases where the biomarker is a measured quantity that correlates with a magnitude of outcome, that quantitative relationship must be accurate and reproducible under the conditions in which the biomarker will be assessed in practice. In cases where the application involves a binary test (the outcome or end point to be indicated either is or is not present), values of sensitivity and specificity close to unity are required for reliability.
ARMY APPLICATION AREAS
Neuroscience represents at once both a challenge and a great opportunity. It is a challenge because the breadth and complexity of contemporary neuroscience are so great; and it is an opportunity because neuroscience can arguably become an important vehicle on which the Army depends to achieve its mission goals.
The formal pursuit of neuroscience is a theoretical endeavor on the one hand and a practical area of application on the other. Imaging technologies that form the basis for advances in neuroscience have their roots in the medical arena, and it is the Army Medical Research and Materiel Command that has traditionally sponsored much of the basic neuroscience research of benefit to individuals in all military services. DOD-level recognition of the importance of neuroscience research can be seen in the 2008 establishment of the Defense Centers of Excellence for Psychological Health and Traumatic Brain Injury under the assistant secretary of defense for health affairs.
The Army Research Institute for the Behavioral and Social Sciences (ARI) has traditionally conducted research in support of personnel testing and assessment, but neither ARI nor the Army’s main research arm, the Army Research Laboratory (ARL), possesses in-house facilities to perform basic neuroscience research. In light of a growing awareness of neuroscience potential in military applications, however, the ARL is planning to establish a collaborative technology alliance on cognition and neuroergonomics, which will take a multidimensional approach (e.g., genetics, computational modeling, neuroimaging, and performance) to optimizing information transfer between the system and the soldier, identifying mental states and individual differences that impact mission-relevant decision making, and developing technologies for individualized analyses of neurally based processing in operational environments.6
Neuroscience advances have already led to a broad array of commercial applications and sparked centers for neuroscience research at academic institutions throughout the country and the world. Table 2-1 lists sample objectives in important Army application areas likely to benefit from neuroscience advances. To respond to the statement of task,
6
Kaleb McDowell, U.S. Army Research Laboratory, “ARL Research in Neuroscience,” presentation to the committee, December 18, 2007.
OCR for page 19
Opportunities in Neuroscience for Future Army Applications
TABLE 2-1 Prospective Army Applications for Neuroscience
Application Areas
Sample Objectives
Training and learning
Training paradigms and methods
Shorten training cycles; assess training effectiveness
Performance assessments of individuals and groups
Detect individual performance degradation; assess group–individual interactions
Identification of training candidates
Improve success rates
Training effectiveness measures
Predict optimal performance; anticipate degraded performance
Optimizing decision making
Individual and unit readiness
Utilize neural-state indicators
Adversary assessment and prediction
Act inside adversary decision cycle; disrupt adversary decision making (psychological operations)
Setting objectives
Reduce risk by matching goals with performance
Sustaining soldier performance
Recovery and reset
Mitigate effect of sleep deprivation on recovery; neuropharmacological intervention to mitigate trauma response
Counterstress
Insulate immune system; moderate disease; modify brain functions to contend with combat rigors
Fatigue and pain
Nutritional countermeasures; minimize effects of sleep deprivation; drug therapies
Brain injury
Intervene early to mitigate acute and long-term deficits due to trauma
Improving cognitive and behavioral performance
Soldier skills
Optimize brain–machine interfaces; improve image interpretation capabilities
Information utilization and management
Personalize data fusion; prevent information overload
the applications are organized in four categories: training and learning, optimizing decision making, sustaining soldier performance, and improving cognitive and behavioral performance. There is no question that neuroscience research has great potential for the Army’s future, but there are societal issues, including ethical considerations and cultural impediments, that must be overcome to realize its full potential.
SOCIETAL ISSUES
The decoding of the human genome and the emergence of new imaging modalities are making possible the identification of proteomic, genomic, and imaging biomarkers associated with susceptibility to a specific disease, with environmental stressors, or with neuropsychological vulnerabilities (e.g., pain, reduced perception, anxiety). The aggregate of multiple biomarkers may provide a susceptibility profile that would not be achievable through testing for any single marker alone. These aggregate data can help in monitoring the rate of progression of clinical disorders or response to treatment. Creating a susceptibility profile of such signatures for a patient can allow for personalized medicine tailored to individual need. The nature (e.g., mutation sites, triplet repeats, proteomic signatures, and structural and functional imaging changes) and quantity of biomarkers involved could play an important role in screening for, diagnosing, and predicting disease. This same capability has made it possible to select persons with a low risk of developing disease or succumbing to a variety of stressors (toxic materials in the environment, for instance). There may be adverse economic consequences (uninsurability, reduced rates of compensation) for the individual and his or her career path progression (costly training programs may intentionally preclude high-risk persons from participation) associated with the identification of disease potential and susceptibility to stress, and so the downside of such information has become of social concern.
Ethical Considerations
One consequence of the genetic screening of large numbers of healthy persons for susceptibility to treatable or manageable disease is that subsequent studies may reveal that the same gene predisposes to an untreatable disease. A case in point is the screening of individuals for a particular allele of apolipoprotein E4, which was known in the 1980s to be associated with high risk for cardiovascular disorder. In the early 1990s it was found that an apolipoprotein E4 also was associated with higher risk for Alzheimer’s disease (Corder et al., 1993). Patients who wanted to reduce the risk of cardiovascular disease and signed a consent form for such analysis became aware of their increased susceptibility to Alzheimer’s, a then untreatable neurological disorder. Such information was not wanted and caused distress for a significant number of people.
Another consequence of learning one’s susceptibility to a nontreatable disease is the still-healthy patient’s inability to anticipate the effect of such information on lifestyle and quality of life. Huntington’s disease is a clear example of this effect. In the early part of the 1980s it became possible to screen patients and determine whether they would develop the neurodegenerative disorder Huntington’s disease, a dementia that does not appear until the fourth decade of life or later and that is associated with an extensive triplet repeat of cytosine-adenine-guanine (CAG) (Myers et al., 1993). Because the disease is autosomal dominant (all individuals with the gene will develop Huntington’s disease), it was
OCR for page 20
Opportunities in Neuroscience for Future Army Applications
proposed that all conceptuses that had a parent or close relative with Huntington’s disease should be tested for the gene. Knowledge of whether one has a genetic predisposition and hence was certain to develop the disease could have a neutral, beneficial, or adverse effect on the patient, with outcomes ranging from acceptance to suicide. Undesired information on susceptibility to an untreatable disease might be a sword of Damocles for a young soldier.
The positive aspects of knowing how one’s genetic heritage impacts wellness and resilience to toxic insult can be illustrated by the case of glutathione S-transferase genotype (GST) and resistance of smokers to lung cancer. A significant number of studies reveal a twofold increase in the incidence of squamous cell carcinoma for patients having GST M1 and GST T1 null genotypes. GST and cytochrome P450 are two classes of enzymes that metabolize and detoxify potential environmental toxins. Patients with the null forms do not express active GST with the properties of GST M1 or GST T1, and the absence of these enzyme variants might predispose them to the toxic effects of various chemicals, including some chemical warfare agents or toxic industrial materials. Restricting the assignment of certain soldiers to areas of high risk might protect them from exposure to such toxic materials; however, it might also keep her/him from serving or being advanced (promoted) in an area of specific interest, or it might prevent him/her from participating in important missions before any clinical manifestation of illness.
Federal laws and regulations contained in the Health Insurance Portability and Accountability Act (HIPAA) protect patients and the community at large against the unwarranted and unnecessary disclosure of medical information that is directly or indirectly traceable to a particular individual to unauthorized parties. The primary concerns are that such disclosure might (1) affect the promotion of military or civilian persons in their field of specialization, (2) affect insurance rates and insurability of a given individual, or (3) affect the psychological/social well-being of individuals with catastrophic diseases that are currently largely untreatable (e.g. Huntington’s disease, Alzheimer’s disease). These are but a few of the unintended consequences of inappropriate disclosure.
For additional information on ethical issues relevant to neuroscience research discussion in this report, the committee recommends the following sources: Karanasiou et al., 2008; Fins and Shapiro, 2007; Illes, 2007a and 2007b.
Cultural Impediments
The emergence of new bio- and neurotechnologies permits categorizing the human population into subsets having either increased or decreased susceptibility to disease and stress on the basis of their genotype and their phenotype.7 There is tension between the idea of “selecting out” individuals for tasks based on presumed genetic susceptibilities and the belief that extensive training can overcome inherited limitations and liabilities. The science-fiction film Gattaca (Columbia Pictures Corporation, 1997) confronts this dilemma with a dark view of the preselection concept. The film was made during the early stages of the Human Genome Project and the first cloning of large animals.
The U.S. military community traditionally aspires to select individuals for particular tasks or promote them based on excellence during training and performance in the field. Selecting in for extended service in closed platforms such as submarines is rigorous: The training periods are long and there are particular social/psychological requirements. Despite this, very little research has been done on selecting in individuals who have a particular aptitude as assessed by genetic and phenotype testing for a particular military position or job. The decoding of the human genome and the advent of real-time imaging of neural information flow and noninvasive tracing of major fiber pathways provide an opportunity to learn how we can use these novel methodologies to enhance training and personalize it to meet the needs of the soldier, to identify characteristics that are particularly well suited to complex and extreme environments, and to detect the early appearance of uncompensated responses to stress and emerging TBI and post-traumatic stress disorder (PTSD).
The issues confronting the Army include training to criterion (90 percent or better appropriate response to challenge), increasing data flow from deployed aerial and ground sensors, human intelligence, electronic communications, and tempo of engagement with increasing capability of lethality. The need to reduce casualties during force-on-force engagement drives the development of means for conducting combat at large standoff distances and acquiring extensive awareness of the adversary’s deployment and capability. At the same time, there is a perceived need to minimize noncombatant casualties, which militates against extensive standoffs. These challenges call for a strategy that allows human cognitive capabilities to operate for 18-20 hours per day, 7 days a week for 12 to 15 months at a high tempo of operations. The most affected group will be the command organization, which is permitted little or no respite from high-tempo decision making and little organized sleep.
The Use and Abuse of Socially Sensitive Demographic Categories as Indicators of Neural State and Performance Capability
As the preceding discussions on societal issues suggest, the committee supports and encourages scientifically validated neuroscience applications across the Army-relevant areas highlighted in Table 2-1 and addressed in detail throughout the report. Ethical considerations, such as those related to genetic screening or improper disclosure of
7
Phenotype is the result of genes plus environment, and epigenetic changes that occur in the individual after conception may play an important role.
OCR for page 21
Opportunities in Neuroscience for Future Army Applications
personal medical information, are important constraints that need to be considered, even when the science is adequate for a potential application. In addition, there are a number of issues that require further consideration before being pursued for possible Army application.
One of these issues—the use of performance-enhancing pharmacological agents—was central to the committee’s decision to distinguish between uses of neuroscience-related countermeasures (including pharmacological agents) to ameliorate a deficit in performance due to a stressor (Chapter 5) and uses of pharmacological agents to enhance performance beyond an individual’s baseline capability (Chapter 6). Cautions and caveats appropriate for these two contexts of application are included in both chapters.
A second major area in which caution must be exercised when considering application to Army-relevant problems concerns statistically significant differences in group-averaged neural states or activity patterns—or even differences in behavior—between a demographically defined subpopulation and a reference population. In the case of gender, such differences are typically expressed as a comparison of male and female subpopulations. For other subpopulations of societal interest, such as ethnocultural identity, age, or socioeconomic status, comparisons may be drawn either between one such category (e.g., African-Americans, young people between 18 and 25) and the general population or among subpopulation categories within a classification (e.g., comparisons among ethnocultural groups or age groups).
In most cases, these differences, even when statistically significant, represent differences in population distributions where the distributions have substantial overlap. From an epidemiological perspective, the statistically significant difference in the distributions justifies identifying membership in certain subpopulation categories as a differential risk factor. However, for purposes such as selecting, assigning, or qualifying individuals for a task, the overlap in the subpopulation distributions means that these categories lack the sensitivity and specificity to be reliable indicators of the neural state or behavior of interest. Rather than relying on a familiar but scientifically indefensible population category as a criterion, the appropriate use of neuroscience insights is to seek out one or more truly reliable indicators for the variable of interest.
As a simple but germane example, a number of behavioral epidemiologic studies have found that women are at greater risk for developing PTSD, given similar stress experiences, than men (e.g., Breslau and Anthony, 2007; Turner et al. 2007). Should this difference in relative risk be used, for example, to exclude women from high-stress combat situations? The committee’s position is that gender is not a sufficiently reliable indicator of the PTSD outcome to be used as a criterion in selecting and assigning individual women, even though the studies establish being female as a risk factor for PTSD. The numbers of false positives and false negatives are too high; the correlation lacks sensitivity and specificity.
The work by Ursano et al. (2008) illustrates how the neurosciences can extend and inform behavior-based findings—such as the PTSD studies cited above—by opening the way to reliable indicators. Their work indicates that the 5-HT2A receptor, p11 protein, and associated regulators may play a role in PTSD-related response to stress experiences. If this still-preliminary line of inquiry were to lead to suitably sensitive and specific indicators of PTSD susceptibility, then those indicators would be candidate criteria for selection and assignment decisions, where gender is not. In short, whether one is female or male is not the issue; it is one’s neurophysiological sensitivity to a definable level and type of environmental stress for which the Army needs validated, reliable indicators.
Over the past half-century and longer, American society has traveled a long and difficult road to break away from unscientific stereotypes about gender, “race,” and other previously accepted ways of categorizing individuals to define their suitability for various roles and responsibilities. The committee’s deliberations on how to deal with research on gender differences and other demographic categories acknowledged these societal issues. In light of its concerns about applying group-averaged statistical differences to individuals within a group, the committee decided to emphasize individual variability in neural-based traits and tendencies as a more appropriate way to address observed distributions in a population of interest. By focusing attention on individual variability and the search for reliable indicators of that variability, the committee hopes to avoid unscientific and unethical application of findings about behavior and neurophysiology in ways that would negate our hard-won progress toward fair treatment and equal opportunity.
REFERENCES
Aston-Jones, G., and J.D. Cohen. 2005a. Adaptive gain and the role of the locus coeruleus-norepinephrine system in optimal performance. Journal of Comparative Neurology 493(1): 99-110.
Aston-Jones, G., and J.D. Cohen. 2005b. An integrative theory of locus coeruleus-norepinephrine function: Adaptive gain and optimal performance. Annual Review of Neuroscience 28: 403-450.
Atkinson, A.J., Jr., W.A. Colburn, V.G. DeGruttola, D.L. DeMets, G.J. Downing, D.F. Hoth, J.A. Oates, C.C. Peck, R.T. Schooley, B.A. Spilker, J. Woodcock, and S.L. Zeger. 2001. Biomarkers and surrogate endpoints: Preferred definitions and conceptual framework. Clinical Pharmacology & Therapeutics 69(3): 89-95.
Bakker, A., C.B. Kirwan, M. Miller, and C.E.L. Stark. 2008. Pattern separation in the human hippocampal CA3 and dentate gyrus. Science 319(5870): 1640-1642.
Belenky, G., N.J. Wesensten, D.R. Thorne, M.L. Thomas, H.C. Sing, D.P. Redmond, M.B. Russo, and T.J. Balkin. 2003. Patterns of performance degradation and restoration during sleep restriction and subsequent recovery: A sleep dose-response study. Journal of Sleep Research 12(1): 1-12.
Breslau, N., and J.C. Anthony. 2007. Gender differences in the sensitivity to posttraumatic stress disorder: An epidemiological study of urban young adults. Journal of Abnormal Psychology 116(3): 607-611.
OCR for page 22
Opportunities in Neuroscience for Future Army Applications
Corder, E.H., A.M. Saunders, W.J. Strittmatter, D.E. Schmechel, P.C. Gaskell, G.W. Small, A.D. Roses, J.L. Haines, and M.A. Pericak-Vance. 1993. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 261(5123): 921-923.
Fins, J.J., and Z.E. Shapiro. 2007. Neuroimaging and neuroethics: Clinical and policy considerations. Current Opinion in Neurology 20(6): 650-654.
Friston, K.J., E. Zarahn, O. Josephs, R.N.A. Henson, and A.M. Dale. 1999. Stochastic designs in event-related fMRI. NeuroImage 10(5): 607-619.
Genik, R.J., C.C. Green, F.X. Graydon, and R.E. Armstrong. 2005. Cognitive avionics and watching spaceflight crews think: Generation-after-next research tools in functional neuroimaging. Aviation, Space, and Environmental Medicine 76(Supplement 1): B208-B212.
Illes, J. 2007a. Empirical neuroethics. Can brain imaging visualize human thought? Why is neuroethics interested in such a possibility? EMBO Reports 8(S1): S57-S60.
Illes, J. 2007b. Neuroethics in a new era of neuroimaging. Pp. 99-103 in Defining Right and Wrong in Brain Science: Essential Readings in Neuroethics. W. Gannon, ed. Washington, D.C.: Dana Press.
Karanasiou, I.S., C.G. Biniaris, and A.J. Marsh. 2008. Ethical issues of brain functional imaging: Reading your mind. Studies in Health Technology and Informatics 137: 310-320.
Kurzweil, R. 2005. The Singularity Is Near. New York, N.Y.: Viking Press.
Myers, R.H., M.E. MacDonald, W.J. Koroshetz, M.P. Duyao, C.M. Ambrose, S.A.M. Taylor, G. Barnes, J. Srinidhi, C.S. Lin, W.L. Whaley, A.M. Lazzarini, M. Schwarz, G. Wolff, E.D. Bird, J.-P.G. Vonsattel, and J.F. Gusella. 1993. De novo expansion of a (CAG)n repeat in sporadic Huntington’s disease. Nature Genetics 5(2): 168-173.
Otten, L.J., R.N.A. Henson, and M.D. Rugg. 2002. State-related and item-related neural correlates of successful memory encoding. Nature Neuroscience 5(12): 1339-1344.
Penfield, W., and P. Perot. 1963. The brain’s record of auditory and visual experience: A final discussion. Brain 86(Part 4): 595-696.
Turner, J.B., N.A. Turse, and B.P. Dohrenwend. 2007. Circumstances of service and gender differences in war-related PTSD: Findings from the National Vietnam Veteran Readjustment Study. Journal of Traumatic Stress 20(4): 643-649.
Ursano, R.J., H. Li, L. Zhang, C.J. Hough, C.S. Fullerton, D.M. Benedek, T.A. Grieger, and H.C. Holloway. 2008. Models of PTSD and traumatic stress: The importance of research “from bedside to bench to bedside.” Progress in Brain Research 167: 203-215.
Items in cart [0]
