Neuroeconomic approaches to mental disorders

Abstract

The pervasiveness of decision-making in every area of human endeavor highlights the importance of understanding choice mechanisms and their detailed relationship to underlying neurobiological function. This review surveys the recent and productive application of game-theoretic probes (economic games) to mental disorders. Such games typically possess concrete concepts of optimal play, thus providing quantitative ways to track when subjects' choices match or deviate from optimal. This feature equips economic games with natural classes of control signals that should guide learning and choice in the agents that play them. These signals and their underlying physical correlates in the brain are now being used to generate objective biomarkers that may prove useful for exposing and understanding the neurogenetic basis of normal and pathological human cognition. Thus, game-theoretic probes represent some of the first steps toward producing computationally principled, objective measures of cognitive function and dysfunction useful for the diagnosis, treatment, and understanding of mental disorders.

Figure 1. Nash Equilibrium

In a strategic interaction between two agents, the Nash equilibria are any of the set of choices made by the two agents where any unilateral change in strategy by one or the other agent does not improve the agent’s outcome. Illustrated are payoff tables for a two-agent game where two agents (players) move simultaneously given two options (actions). (top) The payoff table for any given set of choices made by the two agents. The green circles highlight the Nash equilibria. When the two players choose the same action (upper left or lower left quadrants) then they are in a Nash equilibrium; in this state any unilateral change in choice results in a worse outcome and the choices are no longer in Nash Equilibrium. (bottom) payoff tables show the best choices for player 1 (left) given player 2’s choices and for player 2 given player 1’s choices. The set of players’ choices in this example will tend to maintain Nash Equilibrium (one of the two green circles), once it is discovered, since these provide the best possible outcomes for the pair.

Figure 2. Neuroeconomic approaches to studying choice behavior in non-human primates paired with single unit electrophysiological recordings

(Top) Non-human primates are trained to perform choice tasks using eye movements: (left) monkeys fixate until cued to choose “action 1” or “action 2”, which in this example is look to the red circle or look to the green circle, respectively. The circles placements are chosen based on the receptive field of the single unit under study thus allowing study of the relationship between single unit activity and the choice to take “action 1” or “action 2”. In these studies, the reward is typical a squirt of juice or water which is desirable for water deprived (thirsty) players. (Bottom) three game theoretic settings used in non-human primates to investigate the connection between value, choice, and neural activity: (left) payoff table for a two player Inspection game, (middle) Dynamic foraging task, and (right) Matching pennies game played between a monkey and a computer algorithm. These games and other uses of game theoretic approaches are beginning to expose and model expected changes in subjects’ behavior.

Figure 3. Biomarker for borderline personality disorder exposed by multi-round trust game

(A) Iterated trust game: Investors are endowed with $20 at the start of each of 10 rounds. Investors can invest any portion of the endowment with their partner. The invested amount is tripled before being passed to the trustee, who can repay any portion of the tripled investment. (B) The social norm for investor behavior in trust game can be represented as distributions of likely investments given prior expectations and accrued experience. Investments are typically large in this game of cooperation. Thus, small investments represents a deviation from the social norm of the task. (C) A region-of-interest analysis of anterior insula of 38 healthy trustees shows hemodynamic activity to be negatively related to size of investment, consistent with the idea that small investments represent a deviation in norm. In contrast, similar analyses among 55 individuals with BPD showed no significant activation in these regions when investments were small, suggesting low investments do not represent a violation of an established norm. (D) Healthy trustees are twice as likely as BPD trustees to coax when cooperation between players is low. Specifically, healthy trustees are more likely to make a large repayment (≥ investment amount) after having received a small investment. Conversely, BPD trustees are more likely to make a small repayment (< investment amount) after receiving a small investment.