Interdisciplinary Approaches to Neuroscience

0 downloads 0 Views 317KB Size Report
Sep 30, 2012 - For such understanding it is important to take into account the architectural and ... interdisciplinary tool for the study of cognition may have in diverse fields of Neuroscience ...... [16] Andersen RA, Essick GK, Siegel RM.
Interdisciplinary Approaches Epistemology and Cognition

to

Neuroscience

September 30, 2012 | ISBN-10: 1619422735 | ISBN-13: 978-1619422735 | Edition: 1 Nova Science Publishers Inc/New York - (September 30, 2012)

Editor: Tobias A. Mattei, MD

HTTP://WWW.AMAZON.COM/INTERDISCIPLINARY-APPROACHES-NEUROSCIENCEEPISTEMOLOGYCOGNITION/DP/1619422735/REF=SR_1_1?IE=UTF8&QID=1376603809&SR=81&KEYWORDS=MATTEI+TOBIAS

CHAPTER 5 ‘NEUROECONOMICS-ORIENTED’ PSYCHOPATHOLOGY: A NEW APPROACH FOR THE STUDY OF PSYCHIATRIC SYNDROMES Tobias Alécio Mattei, M.D. Neurosurgery Department – University of Illinois at Peoria/IL – US

2

Tobias Alécio Mattei

ABSTRACT Neuroeconomics is a recent scientific approach which has been proposed in order to describe brain functions as an alternative to the classical sensorimotor model based on a deterministic reflex theory. Neuroeconomics’ analysis is based on the methodologies developed by Ecological Biology and Modern Economics. It frequently utilizes concepts of Bayesian statistics (such as probability and relative utility) to explain behavior elicited by an environmental stimuli. In this essay the authors illustrate how the principles of Neuroeconomics can be successfully applied to clinical psychiatric cases in order to represent them as failure in the process of utility maximization. Two practical examples are given: the first representing ‘persecution delirium’ of schizophrenia as a disturbance in the variable of ‘posterior probability’ and the second correlating apathy of major depressive disorders with a disturb in ‘expected utility’. Throughout this paper the author emphasize the benefits of such an interdisciplinary approach for the study of human behavior pointing out the major directions for future investigations in this promising field between Neuroeconomics and Psychiatry.

INTRODUCTION Human Psychology can be understood as the scientific endevour of studying and explaining human behavior. For such understanding it is important to take into account the architectural and functional aspects of the human brain as well as its correlation with the observed external behavior. Therefore, in order to accurately describe both normal and pathological human behavior, the theoretical concepts of Psychology should be supported by empirical observations about the biological functions of human brain. In this context we propose to analyze the practical implications for Clinical Psychology of the recent discoveries of Neuroeconomics, which demonstrates that the human brain encodes in a very concrete, biological and even synaptically level, economical concepts such as ‘event probability’ and ‘expected utility’. By emphasizing both the behavioral expression of specific neuronal firing patterns as well as the electrophysiological activity of clusters of neurons that mediate decision-making under conditions of uncertainty, the emergent discipline of Neuroeconomics has been considered a powerful toll in order to study, describe and explain the neurobiology and behavior during choice and decision-making. It is still early to predict all the impact such an interdisciplinary tool for the study of cognition may have in diverse fields of Neuroscience and psychology. In this paper the author delineates some implications of such Neuroeconomical approach to Clinical Psychology and the study of pathological behavior present in psychiatric disorders. The authors emphasize the practical consequences and scientific repercussions of the correlation of concepts currently used in game theory and economics (such as event probability and expected utility) with psychopathological entities currently used to describe symptoms of psychiatric disorders (such as apathy and persecution delirium). Finally some of interessant studies on Neuroeconomics are presented and the directions for future investigation in this promising research area are pointed out

Neuroeconomics-Oriented Psychopathology

3

HISTORY OF CLASSICAL DETERMINISM Since Descartes’ proposal of behavior description with basis on deterministic responses to sensorial stimuli, the leading conceptual view in neurophysiology has been the ‘Classical Reflex Theory’. Descartes believed that every behavior, even the most complex ones, could be explained by mechanistic reflex- ike mechanisms. [1] This view was in close correlation with the premises held by the Cartesian Ontology [2], which proposed to understand the world as a spectacular and complex ‘clockwork’, amenable to extensive analytical description by the laws of classical physics. These philosophical ideas were not new and most of them could be found in texts of greek philosophers Epicurus and Democritus, who argued that the world was composed entirely of matter and, therefore, an extensive analysis of the causal interactions among the particles, should in principle, account for all physical events. According to this view if all the events occuring in the universe are the product of the collision of small particles that strictly obey fixed physical laws, the human behavior, should, therefore, also be a product of the interaction among such particles. These deterministic ideas acquired during the Renaissance period a distinct place of honor among philosophers, being considered as the logical consequence of a scientific interpretation of the world. The Scottish philosopher David Hume, for example, [3] suggested that every human and animal action could be reduced to a complex series of deterministic interactions. According to this view, there was a naive hope that the mathematical tools as well as the recent development in classical Physics, would be enough to describe both human behavior and the underlying operation of human brain and mind. A good illustration of such deterministic way of thinking is the famous LaPlace’s idea of a perfect mind. [4] This philosopher argued that if someone dominated all the possible informations about the present world (in other words, if someone knew all the current positions and velocities of each tiny physical particle of the universe) then this person should be capable, at least in theory, to preview all the future states of such particles, and therefore, all future events. The neurophysiological approach based in such deterministic philosophical view of the world was the classical ‘Reflex Theory’, which postulated that sensorial stimuli acting on the nervous system would elicit a motor behavior which could be fully predicted. Brain and analogous structures would be, according to this view, essentially passive and serving as a form of conduct through which impulses and forces could be transmitted from sensory afferents to motor efferents pathways. This view gained significant attention when the eighteen-century english physiologist Charles Scoot Sherrington [5] combined Descartes` philosophical model with a complex logical-analytical system developed by him, forming the core of a deterministic theory for explaining the biological basis of behavior which lasted for more than two centuries. Other important support for the biological basis of behavior came from the works of the Russian physiologist Ivan Pavlov, [6] which suggested that every behavioral pattern was essentially deterministic and that a purely mathematical understanding of sensorial-motor reflexes would be the most adequate system for describing such behavior.

4

Tobias Alécio Mattei

THE BIRTH OF NEUROECONOMICS While deterministic psychophysicists and sensory physiologists traditionally emphasized the effects of sensory stimuli on decision-making, cognitive psychologists and economists have for a long time suggested that decision-making is strongly influenced by the subjects’ prior experience as well as from their beliefs regarding the 'value' or ‘utility’ of alternative choices. Among the great critics of the simplicity of reflex theory of behavior it is worth mentioning the English biologist Grahan Brown and the german physiologist Von Holst. [7] These scientists argued that, at the end of the nineteen-century, there was enough empirical evidence to support the affirmation that the human nervous system had much more complex role than the simple conduction of sensorial stimuli. In fact, they strongly defended that the nervous system could actually generate its own activity apart from any sensorial stimuli. Working from another starting point, a second group of anti-reflexive neurobiologists, composed by scientists such as viennese physiologist Paul Weiss [8] and the russian cyberneticist Nikolai Bernstein, [9] argued that while the reflex seemed to be a good model for many of the simple observed behaviors, whenever a detailed analysis was performed the scientist would find that such behavioral pattern was organized around well-defined goals rather than being a loose conglomeration of local and independent reflexes. Nevertheless, the main contributor for the creation of a robust alternative approach to the classical deterministic reflex theory was the English computer-scientist David Marr. [11] Early in his career Marr seems to have been convinced by the idea that formal computational studies of the nervous system were the best way to achieve a deep understanding of brain functions. In a famous paper published in 1976 Marr and Poggio [12] argued that the scientists should not isolate a tiny piece of behavior and further try to figure out which “specific nervous pathway was responsible for that behavior, ultimately building a theory of the brain function out of these tiny pieces. In opposition, he defended that one should start from a general overview, trying first to understand the computational goal which the neurobiological system and its architecture was attempting to accomplish. In other words, the scientist should begin from the ‘top’ by describing what the whole system was attempting to accomplish (as formally and mathematically as possible) and only then begin to ask how the biological hardware would achieve that goal or computation. In Marr`s own words: “To understand the relationship between behavior and brain one has to begin by defining the function or computational goa, of a complete behavior. Only then can the neuroscientist determine how the brain achieves that goal”. In this attempt to describe the biological systems based on its computational goals, Marr and colleagues found that Economics’ concepts from Statistical Theory such as ‘probability’, ‘expected utility’ and ‘optimal responses’ were much more useful than the classical mathematical tools. These new approach to understand brain through the description based on

Neuroeconomics-Oriented Psychopathology

5

modern economic theory and behavioral ecology was the beginning of a new born theoretical research line called Neuroeconomics. The knowledge in Neuroeconomics has exponentially growed since the 1970s, and recent experimental works have demonstrated that the firing pattern of specific neuronal clusters of the brain may be optimally described by such economical tools. Of special interest is the work of Glimcher and col. about the role of posterior parietal cortex (more specifically the Brodmann area 7, known as LIP area) in visual saccadic movements. [13] [14] These researchers have demonstrated that classical concepts of sensory and motor areas were not capable to give a good explanation to patterns of neuronal activation of area LIP observed in animal experiments. Historically, while Goldberg and col. [15] proposed to describe area LIP as having a role in motor planning, and therefore, an ‘intentional character’, the group of Anderson and col. [16] tried to describe LIP as a sensory association area, with an ‘attentional character’. Nevertheless, neither Goldberg nor Andersen interpretations could predict with precision the results of such exmpirical experiments, which, sometimes tended to favor the first view and other times the second. It seemed that the theoretical framework for interpreting the experiments were failing. In this context Glimcher et col. proposed a new approach to the problem with basis on classical economical concepts. The most interesting fact is that this economical approach was able to describe perfectly all the aspects of the experiments, providing excellent tolls to describe the activation pattern of that specific area of the human brain. In fact, Glimcher’s group demonstrated that neuronal activity in area LIP was intrinsically correlated with two specific economical variables (‘probability’ future event occurrence and the ‘expected utility’ for such occurrence). Other promising line of experiments in Neuroeconomics, which have already demonstrated consistent results, is the study of sensorimotor systems in terms of ‘utility functions’. [25] The theoretical basis for such experiments relies on the fact that every sensorimotor system has to choose between different actions. The practical occurrence of every action will depend basically on two factors — the cost associated with performing the action and the desirability of the produced outcome. Through an analysis of the variability and predictability of these actions a curve of ‘utility function’ of the behavior may be constructed. (Figure 1) These utility functions can be experimentally assessed by measuring the ‘indifference curves’ derived from empirical experiments employing decision-making paradigms. In sensorimotor control the utility functions often depend on several variables. Consider, for example, unpacking a car after a snowboarding vacation. One could carry all the suitcases at the same time, reducing the time to unpack but maximizing the weight he has to lift at one time. At the other extreme one could transport each item individually, which would minimize the magnitude of the force required at the expense of a long unpacking duration. The chosen solution is likely to lie somewhere between these two extremes and may reflect a ‘optimal’ decision based on a utility function that depends on the specific personal preferences of each subject. Once a ‘utility function’ is specified, the decision problem becomes solving an optimal control problem and finding the set of actions that maximizes the utility. A number of studies in the field of optimal sensorimotor control have proposed ‘loss functions’ (the negative of ‘utility functions’) and derived the optimal predicted actions given the proposed ‘loss functions’.

6

Tobias Alécio Mattei

For example, in a pardigm called the ‘minimum jerk model’ [10] [24] Hogan et al. suggested that people minimize the average squared jerk of the hand (third derivative of position) when making ‘reaching movements’. Alternative models have suggested that during such movements people try to minimize the variation of endpoint errors that arise from noise on the motor commands [31] [29]. Although Neuroeconomical tools have been employed in several different sets of behavioral studies, most of them evaluating the role of dopaminergic mesolimbic reward circuits involved in decision-making, up to know there has been no application of such powerful tools to the study of clinical psychiatric syndromes. The author believe that such approach may be very useful, because such dopaminergic mesolimbic reward circuits has been demonstrated to be involved in several pathological behaviors such as addiction, obsessive-compulsive disorder, schizophrenia, pathologic gambling, major depressive disorder, and attention-deficit/ hyperactivity disorder. In the sequence the author provide a brief exposition on how this neuroeconomical approach can be extended to the interpretation of psychological features of psychiatry disorders, providing a mathematical description of such pathological behaviors. Two practical examples are used to demonstrate how psychiatric syndromes can be understood as a failure in the task of maximization of the ‘utility function’, leading to a suboptimal and pathological behavior. Afterwards the author delineates the theoretical basis for future experiments necessary in order to correlate the proposed theoretical framework from economic theory with empirical data.

PSYCHIATRY AND NEUROECONOMICS The neuroeconomical approach is based in the assumption that evolutionary processes have selected advantageous behaviors along the time. The operation of the human nervous system could be, on these principles, understood as the biological result of these evolutionary selection. The goal of the nervous system, in evolutionary terms, could be, on these lines, described as the task of evaluate, select and perform decisions that proportionate the maximum evolutionary benefit to the organism. When dealing with the human brain, however, it is important to clarify that such advantages may not be simple and straight-forward but must take into account the whole social system in which the subject is involved. According the results of previous neuroeconomical studies it is supposed that the human brain reflects in a very concrete and even synaptical level economical concepts usually used in evoluationary ecology to describe the process of decision making such as ‘event probability’ and ‘expected utility’. Therefore the author believe that the classical economical tools of Utility Theory would provide a useful mathematical method for analysis of human behavior and underlying brain activity. If this is correct, then the laws and tools of Utility Theory could be used to combine probabilities of future outcomes with the expected utility of each outcome in order to estimate a value for each choice and thus, predict, both human and animal behavior.

Neuroeconomics-Oriented Psychopathology

7

In a simple way, from an economical point of view, the final goal of human mind can be said to be (in a quite simplistic but still true way) the evaluation and selection the set of decisions which provides the maximal expected utility for each specific situation. The total value of the expected utility of an option I (EXI) can be understood as the total amount of final benefits that the decision for the option I brings. Such value can be mathematically determinated by the multiplication of probability of the occurrence of the event ‘x’ given the decision for the option I (EPx/I) and the utility of such event (EUx/I). For calculating the total expected utility of an event which requires more than one decision (for example the decisions I, II and III), it is necessary to multiply the expected utility (EX) for the final event x (EUx/I,II,III) by the respective event probability of the occurrence of the final event ‘x’ given each decision (EPx/I EPx/II, EPx/III). Mathematically we have: n = number of decisions related to event ‘x’ EXx/n = (EPx/I x EUx/I,II,III) + (EPx/II x EUx/I,II,III) + ... + (EPx/n x EUx/n) In order to illustrate these concepts we provide a practical example. Imagine someone who wants to invest 100 dollars in the stock market. There are two classes of papers. The first class has a probability of 30% of giving rentability 50 dollars in one month and 70% of giving a rentability of 20 dollars in the same period. The second class of papers has a probability of 10% of giving a rentability of 90 dollars in one month and a probability of 90% of giving a rentability of 5 dollars. How could one compute these values in order to discover which class of papers offer the best benefits? One solution is to multiplicate each event (or decision) utility (EU) by its event probability (EP) and calculate which class of papers proportionate the greater expected utility. (EX) Mathematically we have: EX = EU x EP In the first case we would have Expected Utility of event (decision) 1 (EX1) = 30% x 50 + 70% x 20 = 29 In the second case we would have Expected utility of event (decision) 2 (EX2) = 10% x 90 + 90% x 10 = 18 As it can be observed the first class of papers (decision 1) proportionate the greatest expected utility and, therefore, is the best option for a rational business man. As we already mentioned in the case of a set of decisions (n) leading to an (x) event it is necessary to calculate separately the event probability that each decision ‘n’ (EPx/n) would ultimately lead to the expected event ‘x’. Once understood the role of ‘event utility’ and ‘event probability’ in determining the ‘expected utility’ we now focus in some of the first experiments which employed

8

Tobias Alécio Mattei

neuroeconomical tolls in order to describe animal behavior in face of a task of ‘choice and reward’ paradigm.

EYE-MOVEMENT AND REWARD EXPERIMENTS The ‘expected utility’ for any event is, in practice, determined through a statistical analysis which considers the ‘event probability’ of prior similar events in the past. First let’s imagine the following experiment: a monkey is posed in front of a white screen and suddenly a spot of light appears right in the middle of the screen. If the monkey looks to the left, it receive 10 ml of grape fruit juice (which monkeys love) in 50% of the times (outcome ‘a’); in the other 50% they receive 5ml (outcome ‘b’). If the monkey looks to the right it receive 1ml of juice in 90% of the events and 20 ml in 10% of the times. In this simple situation, the calculated expected utility (EX1) for looking to the left (event 1) could be described as the product of the expected utility of the first outcome (EU1a) - which is 10 ml - multiplied by the probability of the outcome ‘a’ ocurrence (EP1a) – which is 50% plus the expected utility of the second outcome (EU1b) - which is 5 ml - multiplied by the possibility of the outcome b occurrence (EP1b) - which is 50%. Mathematically for the event (choice) ‘1’ we have: EX1 = EU1a x EP1a + EU1b x EP1b EX1 = 50% x 10ml + 50% x 5ml = 7,5 ml This would give a total expected utility for event 1 (looking to the left) of 7,5 ml. In the same way the expected utility for looking to the right (event 2) can be described as the product of expected utility of the first outcome (EU2a) - which is 4 ml - multiplied by the probability of occurrence of the first outcome (EP1a) - which is 50% - plus the expected utility for the second outcome (EU1b) - which is 1 ml -multiplied by the probability of occurrence of the second outcome b (EP1b) which is 50% . Mathematically we have: EX2 = EU2a x EP2a + EU2b x EP2b EX2 = 50% x 4ml + 50% x 1ml = 2,5 ml This would give a total expected utility for event 2 (looking to the right) of 1,5 ml. The above-presented example illustrates the way economical tolls could be used to predict the monkey’s behavior in such decision-making reward paradigm experiment if they acted in an optimal way in order to maximize utilization, that is, to receive more juice. The most interesting finding is that these mathematical predictions reflect very close what happens in empirical experiments. If the previously calculated expected utility for looking to left is 7,5 and to right is 2,5, the animal submitted to such task, choose in about 75% of the times the first alternative (with the expected utility of 7,5 ml) and in 25% the second alternative (with expected utility with 2,5 ml).

Neuroeconomics-Oriented Psychopathology

9

In an unconscious (and still not yet fully comprehended way) it seems that the brain apparatus of the monkey is able to encod the expected utility of each event and to determinates the best choices in order to maximize this utility. In fact, the authors of such experiments [13] have not only demonstrated that behavior the monkey may be optimally described in economical terms such as ‘expected utility’, but also showed that the electrical activity of some clusters of neurons in the cerebral region responsible for intentional movements presents firing rates which follows exactly such 7,5/2,5 proportion. In other words, researchers were able to demonstrate that there is a biological substrate that encodes the theoretically predicted ‘expected utility’. In other words such ‘expected utility’ values that were first theoretically predicted with economical tools, were successfully confirmed by empirical experiments and even neurophysiologically recorded as a frequency of firing of neuronal cluster cells in the cerebral cortex. These results demonstrated that animal sensorimotor behavior could be described as an optimal economical task of choosing between two alternatives based in their final ‘expected utility’. They also demonstrated that it is possible to correlate the firing patterns of some cluster of neurons in cerebral cortex of the monkeys with the calculated ‘expected utility’ values. It is certainly surprising that the neuroeconomical tolls were useful not only useful to predict the behavior patterns in a choice task, but also to describe the neurophysiology of the animal’s brain involved in such tasks. Based in previous clinical and research activities with the psychopathology of psychiatric diseases, the author developed a framework employing such neuroeconomical tools in order to describe and predict pathological human behavior. In order to prove the usefulness of such approach the author propose two practical examples of how the psychopathological features of psychiatric disorders (the apathy of major depressive disorders and the persecution delirium of schizophrenia) can be mathematically described using same tolls of ‘expected utility’ and ‘event probability’ previously exposed. More specifically such analysis demonstrates how persecution delirium can be understood as a disruption in the optimal calculation of event probability, and how depressive apathy may be represented as a case of a genral depreciation of the expected values for any event’s utility. Our purpose in this introductory essay is not to empirically demonstrate such correlations, but to provide a ‘proof-of-concept’ about the usefulness of employing such economical tools in the psychopathological research of Clinical Psychiatric. In fact, because most of the psychiatric behaviors involve a disruption in the aforementioned dopaminergic mesolimbic reward circuits involved in decision-making, we believe that it is possible to understand as well as represent several other psychiatric symptoms as a disruption in optimal behavior predicted by the Expected-value theory. The key components of such research program is similar to those of previous neuroeconomical studies: First, it is necessary to demonstrate how the pathological behavior is under the control of value computations emerging from an unconscious human effort in order to evaluate the sensorial data and to perform choices which maximize utility. In a second stage, by studying the pathological behavior of real patients, a general model of

10

Tobias Alécio Mattei

behavior for each psychiatric syndromes should be constructed, providing further insight into the decision variables which might influence each set of behavior. Finally it is necessary to correlate the predicted utility functions with functional data obtained by electrophysiological experiments employing techniques such functional Magnectic Resonance (fMRI) and Positron Emission Tomography (PET) scan. In fact, according to previous studies, the measured activity of neuronal clusters evaluated through these functional tools, should be able to confirm the validity of the variables that were taken into account at the time of the construction of the theoretical framework for that specific psychiatric symptom. [22]

PERSECUTION DELIRIUM AND THE BAYES THEOREM Bayes theorem is the statistical method used to describe the possibility of some event occurrence with basis on probability of occurrence of another given event. [17] In practical terms, Bayes theorem provides a value for the probability of occurrence of some event ‘w’ given the fact that another event ‘y’ already occured. Such event probability can be represented as P(w/y). The probability of occurrence of ‘w’ given that ‘y’ has occurred - P(w/y) - can be calculated by multiplying the probability of occurrence of ‘y’ given that ‘w’ occurred - P(y/w) plus the probability of occurrence of w (Pw), divided by the probability of occurrence of y (Py). Mathematically we have (eq. 1):

(eq. 1) According to our proposed thesis the Bayes’ theorem, a statistical toll of Economic Mathematics, could be useful in order to describe the persecution delirium present in schizophrenic syndromes. Persecution delirium is a form of ‘delusion’. Classically a delusion has been defined as a false belief about reality. [13] Such beliefs are usually held by patients with a unusual conviction. Furthermore the absurdity of them are usually clear at the eyes of others and, sometimes, even to the patient himself, who cannot provide a logical explanation for such belief. People with persecutory delirium selectively attend to threatening informations and jump to conclusions on the basis of insufficient information, usually attributing negative events to external personal causes, and misinterpretating others’ intentions, motivations and states of mind. In the sequence we propose a mathematical description in terms of probability theory which could afford for a general description of a persecutory delirium. Let’s suppose the event ‘w’ as being the situation in which the subject is really under persecuted by some secret governmental agents and really in danger, and ‘y’ the clues or evidences (or the set of clues – y1, y2, y3…) which makes the subject believe that he is being persecuted (for example the fact that someone is walking behind him on the street).

Neuroeconomics-Oriented Psychopathology

11

We know “a priori” that at the internal, subjective evaluation of psychotic patients, the probability of being considered under persecution once someone is observed behind them on the street (P(w/y) ) is extremely high. In other words if a psychotic patient with persecution delirium sees someone walking behind him on the street (event y) the patient will have a high suspicious that he is being persecuted (event w). In such psychiatric patients it seems that the brain is acting sub-optimally in the task of evaluating the probabilities for the occurrence of such event. Such pathological behavior, therefore, provides the patient a strong but unreal sensation of being persecuted and, therefore, elicite reactive afraid behavior which is not appropriate for the real circumstance. Following the Bayes Theorem presented above, if the event probability of being persecuted (event w) given that someone is walking behind him on the street (event y) - P(w/y) is high, then, it follows that one of the members of the second part of equation 1 should also be disturbed. Which one would best describe such clinical scenario? Let analyze initially the first term of the second half of the equation, the probability of the occurrence of ‘y’ given that ‘w’ occurred (P(y/w)), which describes the probability of the presence of the cue (y) once someone is really being persecuted (event w). Let’s suppose, for example, a value 50% to P(y/w). In such situation we are supposing that in 50% of the real persecutions a person walking behind someone can be really found to a ‘persecutor’ (we will name them “back-persecutors”). In the remaining 50% of persecutions cases, there is no one behind the person, either because the persecutor walking in the opposite direction along the street, or maybe just observing the victim from a distant point (we will call them “non-back persecutors”). Could we suppose that an elevation of P(y/w) leading to a high probability of being persecuted P(w/y) is the best analytical description of the problem faced by the patient? In this case psychotic patients feeling of being persecuted would be generated by an abnormal high values of (Py/w) – the probability of back-persecutors in comparison to the mean value presented in the general population (which we supposed to be a half to half proportion). In other words, does a schizophrenic patient have such an incontrollable and firm conviction that all persecutors are always behind them and that no one walking in the opposite direction or observing from a distance could ever be a persecutor? This does not seem to be the case. Although schizophrenia seems to lead to several types false beliefs this kind of illogical thoughts does not seem to be the rule observed at the clinical psychiatric practice. From the analysis patients’ behavior and discourse it seems more rational to suppose that the mistakes in such patients’ beliefs are not related to the location from where persecutors come, but to the capacity in differentiation normal followers from persecutors. Let’s, thus, examine the other terms of the equation 1. P(y) in this example represents the probability of finding someone walking behind the subject on the street. We can say that this term has a relative high value, especially in the main streets of the great cities. Note that the term P(y) is in the denominator, and therefore in order to expect a high P(w/y) - which is present the symptom of persecution delirium- the P(y) - the probability of finding someone walking behind - would have to be lower in schizophrenic patients than in the normal population. It also does not seem to be the case that every schizophrenic patients believes that given a low P(y) (in other words, that no one ever walks behind them) therefore, if someone is at a specific time behind them, then, there is a great evidence that they are being persecuted. In fact P(y)

12

Tobias Alécio Mattei

(the probability of finding someone walking behind on the street) seems to be much more directly related to the capacity of observing of actual occurring facts, which is usually not impaired in schizophrenic patients. The only remaining term of equation 1 which could explain such disruption is P(w) - the probability of being persecuted. In fact although the patient’s probability of being persecuted P(w) is similar to that of the normal population, the psychiatric patient’s mind is incapable to judge and evaluate that information in accordance with reality. In the patient’s analysis - as a result of his unfounded beliefs - the general probability of being persecuted is very high. In opposition, for normal people P(w) usually remains very low, unless the person might expect for any other reason some kind of persecution (for example, by having committed a crime). This finally lead us to the conclusion that the high values of P(w/y) in persecution delirium of psychiatric syndromes are consequence of a unexpected high value of P(w), while P(y/w) and P(y) seems not to be unaffected by this disease. As previous described it has already been demonstrated that the ‘expected utility’ for visual saccades are biologically encoded by specific cluster of neurons in area LIP. We could therefore, with good amount of empirical data, suppose that other forms of expect expected utilities may also be biologically encoded. In fact, other authors, such as Hemsley and Garety, [26] have already demonstrated that Bayes’ decision-making theorem can be used as a good theoretical basis for examining reasoning biases in deluded subjects. [27] Furthermore, studies from cognitive neuroscience suggested that hyperactivation of the amygdaloid nucleus (an important part of the limbic system) is associated with fear-like and anxiety components of schizophrenic patients. Excessive activation of such area in these patients was also associated with hypervigillance, particularly toward threat-related cues. It has also been demonstrated that attention to threatening material relevant to the self also differentially activates a more dorsal region on the left inferior frontal gyrus (BA 44). Finally neurophysiological studies with functional magnetic resonance imaging (fMRI) have already demonstrated that people with persecutory delusions regard ambiguous data in the social domain as self-relevant and selectively attend to such threatening information. [21] In such studies while determining self-relevance, deluded subjects demonstrated a marked absence of rostral–ventral anterior cingulate activation activity together with increased posterior cingulate gyrus activation in comparison to the normal subjects. The abnormalities of cingulate gyrus activation while determining ‘self-relevance’ suggest that an impaired self-reflection is present in persecutory deluded states, which may contribute to persecutory belief formation and maintenance. A positron emission tomography (PET) study of a group of patients with chronic schizophrenia [33] demonstrated significant positive correlations between the reality distortion dimension (as measured by validated questionnaires) and the regional blood flow (rCBF) in the left prefrontal cortex and ventral striate areas as well as in the left superior temporal and parahippocampal gyrus. A Single Photon Emission Computed Tomography (SPECT) study of unmedicated patients with schizophrenia [34] also suggested a strong positive correlation between the reality distortion dimension and the left striated area activation as well as a strong negative correlation with left temporal areas.

Neuroeconomics-Oriented Psychopathology

13

All these data highlight the importance of the mesial temporal and ventral striate limbic areas in delusion formation. With basis on such data, future experiments in Neuroconomics of psychiatric diseases should focus on the activation patterns of of such areas. Further correlation of the data obtained by functional experiments with the variables predicted by neuroeconomical tools would lead to a confirmation about the correct theoretical basis proposed for such symptoms. In fact being able to correlate between measurements of brain activity in some specific areas in such patients with predicted abnormal curves of event probability for perception of dangerous events would provide an important advance in study of persecution delirium in schizophrenic patients. A simple empirical paradigm for testing such hypothesis could be, for example, an experiment in which both patients and a control group, while under functional monitoring (for example with fMRI, PET or SPECT) are presented to a movie in which there is someone walking behind a person in the street. The patients are the, asked to imagine the person walking on the street as being themselves. Finally the patients are asked to judge if the person presented in the movie is being persecuted or not. The aforemetnioned theoretical predictions of Neuroeconomics would expect that schizophrenic patient would classify the person of the movie as a persecutor more often than normal subjects. The value and percentages obtained for such answers should be, the, in a second stage, correlated to validated psychiatric scales about the burden of the disease in each patient as well as with the activation patterns observed in the functional experiments. In addition, through a neuroeconomical approach one could hypothesize several other the questions such as: -

-

-

Could the hypervigilance be described by a classical probabilistic function P(w) of an event w? What are the necessary and sufficient characteristics related to this threateningrelated cues which define an event as belonging to the group w? (In mathematical terms it would correspond to define the ‘domain’ and ‘image’ of the function P(w). What is the graphical representation of such probabilistic function? How such functions behave in relation to other variables involved in the Bayesian theorem (such as the previous occurrence of similar events or the presence of distracting variables). What is the temporal evolution of such probabilistic function during progression of the disease and how the use of anti-psychotic medications modifies such mathematical functions?

Many other questions could be formulated, and we believe that such mathematical and conceptual analysis in neuroeconomical terms may provide real and consistent benefits in the scientific effort to describe, understand and represent the particularities and nuances of the complex mental phenomena involved in the psychiatric disorders. In the sequence, we present a second example of how neuroeconomical tools could help Clinical Psychiatry to better understand the pathological behavior observed in psychiatric syndromes. At this time, we will focus on a very important symptom of some psychiatric syndromes: ‘apathy’. Although ‘apathy’ has been classically associated with major depressive disorder, it

14

Tobias Alécio Mattei

is also an important component of several other psychiatric as well as neurological syndromes, such as Alzheimer dementia and Huntington disease.

APATHY AND EXPECTED-VALUE THEORY The simplest definition of apathy is ‘lack of motivation manifested in the absence or reduction in purposeful behavior’. Recently, specific criteria for diagnosing the apathic syndromes as well as scales for evaluation of apathy as a continuous variable have been proposed. The Apathy Evaluation Scale for example, has been demonstrated to provide reliable and validated measuraments of diminished motivation in several clinical populations. [21] The neural mechanisms of apathy are postulated to involve brainstem and forebrain circuits that mediate goal-directed behavior. The functions of these circuits provide a model for classification of apathic syndromes into cognitive, sensory, motor, and affective subtypes. Nevertheless, apart their particularities, eachof these subtypes seems to have a common component, which can be understood in terms of either lack of attention or as a reduction in goal-directed cognitive and cognitive behavior. Although attention itself seems to be more affected by other psychiatric syndromes, such as attention-deficit hyperactive disorder (ADHD), it can also be understood as an important component of apathic syndromes if taken into account the nineteenth-century psychologist William James’ definition: [18] “Attention is that process by which the speed or accuracy of normal sensory processing is enhanced; an enhancement that leads to increase of focus in the sensory world and may be manifested as new active behavior”. Would it be possible to describe such clinical picture present in apathic syndromes employing neuroeconomical tolls? For over a century experimental economists have characterized the decisions people make based on the concept of ‘utility functions’. These functions increase whenever the desirability of the outcome increases, and decrease with the increasing costs associated with performance of the tasks that generate the outcome. For classical economic theory, people make decisions which maximize utility. In other words, every choice is based on an evaluation by the subject on how is it possible to obtain the maximum of benefits with the lowest possible cost. When the ‘utility function’ depends on several variables, ‘indifference curves’ are used to represent those outcomes which are equally desirable (those with similar utility). Whereas in Economics ‘utility’ is studied in terms of goods and services, the sensorimotor system may also be understood as having utility functions defined as the desirability of various behavioral outcomes. [25] As we saw before the goal of human behavior can be understood as the maximization of the the expected utility (EX), which can be mathematically represented as the result of the multiplication of event probability (EP) and event utility (EU): EX = EP x EU

(eq. 2)

Neuroeconomics-Oriented Psychopathology

15

Attention, therefore, can be economically understood as a resource, which may be allocated to different external objects. According to this view, at each moment the person faces the task of distributing her attention according to the expected utility for each option. The more desirable the result of an event (the greater the expected utility) the more attention the person will directed toward such event. In the same way, the higher the associated cost of such task (for example, reading a difficult book) the lower the value of expected utility. We can suppose, then, that attention might have a linear correlation with expected utility. The degree of this correlation in a particular situation will depend also on other factors, such as the subject’s beliefs and mood. However, in a simple neuroeconomical model, the higher the expected utility of some action, the more attention that action will elicit in the organism resulting in allocation of more cognitive resources (perceptive or motor) in order to execute the task. Although attention depends also on other variables (such as the ability of the subject in performing such task – so that new tasks demand more attention that old ones) we may, suppose, initially suppose a linear correlation between ‘attention’ and ‘expected utility’. It is important to emphasize that when considering attention as represented by the ‘expected value theory’ it is also important to take into account the event probability of the outcome (the probability of the occurrence of the outcome once the necessary attentional resources are allocated). This is the reason why even an action with a very high event utility may present with a low final expected utility if the probability of occurrence of such event is low even if all the attentional resources are allocated. Gambling would be a good illustration for that. In suchcase, although the event utility is very high – the costs for performing the action is low (just some cents) and the outcome is highly desirable - the final expected utility of the event is (EX) still low, reason why the majority of the individuals do not do frequently gamble. This occurs because even allocated the necessary resources for the action (in this case the money for the ticket) the probability of occurrence of the desirable event (EP) is still extremely low. One can perceive, therefore, that both event probability (EP) and event utility (EU) are important factors that determine the final expected utility (EX) and, consequently, the selection of a specific behaviour. As already mentioned the relations between these variables may be much more complex if attention is supposed to be a resource that can be represented by continuous instead of allor-nothing variable. Once understood the role of event probability and event utility as well as their relation to the final expected utility and the occurrence of an action, we will now turn our attention back to the psychological symptom of apathy and try to define it in neuroeconomical terms. We could suppose that, for apathic patients, the neural mechanisms underlying the process responsible for evaluation of utility of future events (as well as for selection and allocation of resources in order to accomplish such goals) seems to be somehow disturbed. In such patients it seems that few internal or external factors are capable of eliciting the patient’s attention. In other words, the patient is unable to use his capacity of predicting future benefits of an event in order to allocate his attention and direct his motor behavior toward the accomplishment of such goal. We could say that, in such patients, the expected utility equation must be somehow disturbed. But would it possible to analyze such disturb in more detailed and mathematical

16

Tobias Alécio Mattei

manner? Similarly to previous experiments, would it be possible not only to describe such symptom in economical tools but also to correlate such mathematical data with specific patterns of activation of specific clusters of neurons? In the case of apathy it seems that not only expected utility (EX) for some events, but expected utility as a whole seems to be reduced. This leads us to one of two possible situations: either event probability (EP) or event utility (EU) for events should be reduced. If the event probability of some event is low, the subjective feeling would be something like finding an event very interessant, but judging its occurrence as almost impossible. As already mentioned such (EP) should be low not only for some events, but for every event. However it is difficult to imagine which would be the clinical manifestation of a general reduction in EP for all events. Would the person have an unusual feeling that nothing would ever occur? That sounds very strange and does not seem to correlate with the subjective feelings reported by apathic patients. Additionally, it is difficult to correlate such hypothetical low EP with a disturb in any specific known neuronal cluster of human brain because, up to now, no brain area has been shown to play a role in ‘general representation’ of events’ occurrence. In the second option, in order to have a low expected utility (EX), it would be necessary to have a low event utility (EU). As we saw, in the apathic cases, the event utility of every event should be low. In fact, this seems to correlate very well with sensation experienced by aphatic patients. If the expected utility for all possible events is low, the patient should be expected to present no special interest for any event. In such situation no attention would be elicited by any sensorial event, and, therefore, there would be a lack of intentional and planned motor actions. This picture seems to accurately correspond to the behavior observed in apathic patients. Because expected utility can be mathematically represented by a ‘utility function’, we could demonstrate the following relation: A = lim EX (xn)

0

Being A = apathy EX (xn) = the generic utility function for nth event x After constructing a theoretical basis for predicting and modeling the patient’s behavior in terms of neuroeconomical tools, the future research on apathy should focus on forebrain reward systems which mediates the linking between the future of prediction of a pleasant outcome and the activation of premotor areas responsible for the planning and execution of actions necessary to bring the desired outcome. From an experimental neuroscientific point of view, it is well known that the circuitry of motivated behavior involves a combination of behavioral-specific regions in the hypothalamus as well as a general ‘reward-system’ involving the midbrain, forebrain. Furthermore some catecholaminergic pathways (particularly the mesolimbic dopaminergic system) are known to play a central role in modulating motivated behaviors. [32] Other neuronal system deeply involved in this decision-making process is the medial prefrontal cortex, particularly the anterior paracingulate cortex. The action of the pre-frontal cortex can be understood as to maintain the representation of goals and the means to acquire them. [30]

Neuroeconomics-Oriented Psychopathology

17

It is already know from behavioral neuroscience experiments that patients with depression present decreased bilateral or predominantly ‘left-sided’ activation of medial prefrontal cortex. [19] Other behavioral experiments also suggested that anticipation of increasing monetary gains activates a subcortical region of the ventral striatum. Such activation is accompanied by feelings involving ‘arousal’ and ‘positive valence’. [20] [28] Additionally, some further questions could be proposed in order to investigate phaty from a neuroeconomical point of view: −

− − − −

Is it possible to quantify the magnitude of changes in the expected utility equation observed in psychiatric patients with apathy when compared to normal healthy subjects? Could these modifications be described in a graphic of utility function? Are there any changes in the utility function which might be observed after treatment with antidepressive drugs? Which other variables (for example the presence of depressive thoughts) could possibly interact with such apathy ‘utility function’? Could these other variables also be described in neuroeconomical terms? (Some authors have demonstrated, for example, that there is a raise in amygdaloid nucleus activation during depressive disease which correlates with an increase in frequency and intensity of negative feelings) [23]

The challenging endeavour of empirical experiments in Neuroeconomics has already begun for the research of visual-attentional and motor performance. Nevertheless we still need to extend such efforts in order to investigate other important cognitive functions of the human brain as well as pathological behavior. As previously mentioned, the intention of this introductory essay is not to provide a complete theory of a ‘Neuroeconomical-based’ psychiatric research, but to emphasize the benefits for both Psychology and Clinical Psychiatric of this recent and promising approach to describe, model and understand human behavior. By providing some practical examples as well as the essential steps for successfully applying the tools of neuroeconomics in order to describe common pathologies of Clinical Psychiatry, we hope to foster the attention of the scientific community as well as to collaborate with the further development of future research programs in this field.

CONCLUSION Neuroeconomics provides conceptual tools that have proved to be of great value for analysis of brain and behavior. This essay demonstrates how this promising scientific approach could be useful not only to describe normal behavior, but also to the study of pathological psychiatric syndromes. Throughout this article the author provides a brief overview about the birth of Neuroeconomics as an alternative theoretical approach to the classic reflex theory.

18

Tobias Alécio Mattei

The author also presented and discussed some of the first experiments in Neuroeconomics, related to the description of animal behavior in terms of the ExpectedValue Theory during visual saccadic movements and sensorimotor actions. Finally, the author exemplifies how neuroeconomical tolls could be successfully used in order to describe mathematically two common psychopathological entities of psychiatric disorders (the persecution delirium of schizophrenia and the apathic states of patients with major depressive disorder). We sincerely hope that our short contribution might both present this novel and interesting field to the general scientific community as well as to encourage future theoretical and empirical research in this area, providing the first steps towards the formulation of a ‘Neuroeconomical-oriented’ Psychopathology.

REFERENCES [1] [2] [3] [4] [5] [6] [7]

[8] [9]

[10] [11] [12] [13] [14] [15]

Descartes, R. Passions de l´Ame: The Passion of the Soul, S. Voss (trans.). Indianapolis: Hacket, 1649. Descartes, R. L’Homme - Treatise on Man, T. S. Hall (trans.), Cambridge: Harvard University Press, 1972. Hume, D. Enquiry Concerning Human Understanding. Oxford: Clarendon Press, 1748. LaPlace P.S. Exposition du Système du Monde, Paris: Fayard. 1796. Sherrington, C.S. The Integrative Action of the Nervous System. New York: Charles Scribner´s Son, 1948. Pavlov, L.P Conditioned Reflexes: An Investigation of the Physiological Activity of the Cerebral Cortex, G.V. Anrep (trans.) London: Oxford University Press, 1927. Holst, EV The nature of order in the nervous system. In: Martin RD (transl). Selected Papers of E. von Holst: The Behavioral Physiology of Animals and Man. Miami: University of Miami Press, 1973. Weiss, P. Self-differentiation of the basic pattern of coordination. Comparative Psychology Monographs, 1941. 17(4). Bernstein, N. Trends and problems in the study of physiology of activity. Questions on Philosophy. In Human Motor Actions: Berstein Reasessed, H.T.A. Whiting (ed.) Amsterdam: North-Holland, 1984. Hogan, N. An organizing principle for a class of voluntary movements. J. Neurosci., 1984; 4, 2745–2754. Marr, D. Vision. San Francisco: W. H. Freeman, 1982. Marr, D. and Poggio, T. Cooperative computations of stereo disparity. Science, 1976; 194, 283-287. Glimcher, P.W. Decisions, Uncertainty and the Brain – The Science of Neuroeconomics. Cambridge: MIT Press, 2003. Glimcher, P.W. Eye movement: In M. J. Zigmond et al. (eds). Fundamental Neuroscience. New York: Academic Press, 1999. Goldberg, J.L, Shadlen M.N Representations of perceptual decisions in developing oculomotor commands. Nature. 2000; 404, 390-394.

Neuroeconomics-Oriented Psychopathology

19

[16] Andersen RA, Essick GK, Siegel RM. Neurons of area 7 activated by both visual stimuli and oculomotor behavior. Experimental Brain Research. 1987; 67, 316-322. [17] Bayes, T. An essay toward solving a problem in the doctrine of chances. Philosophical Transactions of Royal Society. 1763; 53, 370-418. [18] James, W. The Principles of Psychology. New York: Henry Holt, 1890. [19] Davidson, R.J. Anterior Electrophysiological asymmetries, emotion and depression: conceptual and methodological conundrums. Psychophysiology. 1998; 35, 607-14. [20] George, M.S, Ketter, T.A, Post, R.M. Prefrontal cortex dysfunction in clinical depression. Depression, 1994; 2, 59-72. [21] Marin, R.S. Apathy: Concept, Syndrome, Neural Mechanisms, and Treatment. Semin. Clin. Neuropsychiatry. 1996; 1(4), 304-314. [22] Leo, P.S, Greg, S.C, William, T.N Choosing the greater of two goods: neural currencies for valuation and decision making. Nature. Reviews Neuroscience, 2005; 6, 363-375. [23] Brian, K. and Richard, P. Neurally Reconstructing Expected Utility, Games and Economic Behavior. 2005; 52(2), 305-315. [24] Flash, T. and Hogan, N. The coordination of arm movements: An experimentally con.rmed mathematical model. J. Neurosci. 1985; 5, 1688–1703. [25] Körding KP, Fukunaga I, Howard IS, Ingram JN, Wolpert DM. A neuroeconomics approach to inferring utility functions in sensorimotor control. PLoS Biol. 2004 Oct;2(10). [26] Hemsley, D.R, Garety, P.A. Formation and maintenance of delusions: a Bayesian analysis. Br. J. Psychiatry. 1986; 149, 51–56. [27] Fischoff, B. and Beyth-Marom, R. Hypothesis evaluation from a Bayesian perspective. Psychol. Rev. 1983; 90, 239–260. [28] Pizzagalli, D.A, Oakes, T.R, Fox, A.S, Chung, M.K, Larson, C.L, Abercrombie, H.C. Functional but not structural subgenual prefrontal cortex abnormalities in melancholia. Mol. Psychiatry. 2004; 9(4), 325, 393-405. [29] Todorov E, Jordan M.I. (2005). Optimal feedback control as a theory of motor coordination. Nat. Neurosci., 5, 1226–1235. [30] Harris C.M, Wolpert, D.M. Signal-dependent noise determines motor planning. Nature. 1998; 394, 780–784. [31] Henrik W, Birgit A, Angela C, Susanne E. Motivating forces of human actions Neuroimaging reward and social interaction. Brain Research Bulletin. 2005 67(5), 368381. [32] McAllister T.W. Apathy. Semin. Clin. Neuropsychiatry. 2000; 5(4), 275-82. [33] Liddle PF, Friston KJ, Frith CD, Hirsch SR, Jones T, Frackowiak RS. Patterns of cerebral blood flow in schizophrenia. Br. J. Psychiatry. 1992 Feb;160:179-86. [34] Parellada E, Catafau AM, Bernardo M, Lomeña F, Catarineu S, González-Monclús E. The resting and activation issue of hypofrontality: a single photon emission computed tomography study in neuroleptic-naive and neuroleptic-free schizophrenic female patients. Biol. Psychiatry. 1998 Oct 15;44(8):787-90.