Neuroanatomy of Decision-making
Human social interactions are extremely complex, and decisions can be influenced by factors such as adaptive strategies, personal preferences, reward evaluation, reinforcement learning, social cooperation, competition and control, as well as other parameters, including uncertainty, ambiguity and probability.[20,48–50] Given the complexity of decision-making, the fact that no single convergent model has yet been proposed is, perhaps, not surprising. Nevertheless, a consensus has been reached concerning a number of fundamental aspects of decision-making. For example, normal decision-making is thought to require an extended neural network, mainly comprising the frontostriatal and limbic loops including serotonergic and dopaminergic pathways, the lateral, medial and orbitofrontal cortex, the striatum, amygdala, basal ganglia, and anterior cingulate cortex.[20,48,49,51] Furthermore, the prefrontal cortex seems to have a critical role in reinforcement-guided decision-making.[52–54]
Three main systems have been suggested to be involved in decision-making: a 'stimulus encoding system', an 'action selection system' and an 'expected reward system' (Figure 1). The stimulus encoding system is important during the evaluation stages of decision-making, and this initial stage seems to be strongly associated with activity in the ventromedial prefrontal cortex, striatum and orbitofrontal cortex,[54,55] and in dopaminergic pathways involving the ventral tegmental area, nucleus accumbens, striatum, and frontal cortex.
A neuroanatomical model of decision-making. Three main systems are thought to be involved in decision-making: a stimulus encoding system (orbitofrontal cortex shown in red), an action selection system (anterior cingulate cortex shown in green) and an expected reward system (basal ganglia and amygdala shown in blue). Other brain areas that are involved in decision-making include the ventromedial prefrontal cortex (stimulus encoding), the lateral prefrontal and parietal cortices (action selection), and the insula (expected reward).
The action selection system is involved in learning and subjective value encoding. Actions that follow a decision seem to be processed predominantly in the anterior cingulate cortex.[55,56] Related processes, such as error perception, as well as exploratory actions and voluntary choices, are also processed in the anterior cingulate cortex. The lateral prefrontal cortex and lateral and medial intraparietal cotices are also activated during this stage of the decision-making process.
The expected reward system is associated with activity in the amygdala, insula cortex, basal ganglia—including the caudate nucleus, putamen and globus pallidus—and the orbitofrontal cortex.[48,57] Reward-based decision-making can be affected by the brain reward system, which consists of the ventral tegmental area, ventral striatum, prefrontal cortex and amygdala. The amygdala also seems to have an important role in emotional learning, which can affect decision-making. The dopaminergic system is thought to modulate the expected reward system, as activity of this system relates to reward learning and prediction of errors.[48,56] In fact, learning the subjective value of objects is critically dependent on midbrain dopamine levels.[49,53]
In addition to these three components, other brain areas are thought to influence decision-making. For example, 'social' paradigms of decision-making—tasks that test an individual's ability to consider the preferences or choices of others—seem to require functioning of brain areas that are associated with theory of mind (ToM), such as the paracingulate cortex, in addition to the striatum, insula cortex, and orbitofrontal cortex.[48,49] Moreover, 'special' forms of decision-making, such as moral decision-making, seem to be associated with activity in the orbitofrontal and dorsolateral prefrontal cortex, cingulate cortex, precuneus and temporoparietal junction. Furthermore, when complex decisions are being made, activity in the insula and posterior cingulate cortex is known to modulate the activity of the prefrontal cortex, and simple tasks involving human volition can engage presupplementary and parietal areas during decision-making. Notably, the numerous brain areas mentioned above interact during different processes. For example, the orbitofrontal cortex is involved not only in evaluation of stimuli but also in prediction of rewards. Similarly, brain areas not classically related to decision-making under ambiguity—the cingulate and parietal cortices—seem to be required for successful performance on the IGT.[23,60] For instance, some patients with lesions in the anterior cingulate cortex might perform normally in the Stroop and Go/No-go tasks, which require action selection—participants must choose between competing stimuli—yet show impaired performance on tasks of decision-making. In summary, decision-making requires coordinated activity within a wide variety of brain areas and neural networks.
© 2010 Nature Publishing Group
Cite this: Decision-making Cognition in Neurodegenerative Diseases - Medscape - Nov 01, 2010.