The role of reward and the basal ganglia in decision making

• Grant number: 13/10694-0
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• Start date: September 01, 2013
• End date: August 31, 2016
• Field of knowledge: Biological Sciences - Physiology - Physiology of Organs and Systems
• Principal Investigator: Nestor Felipe Caticha Alfonso
• Grantee: Carolina Feher da Silva
• Host Institution: Instituto de Física (IF). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

In repeated binary choice tasks, the subject must choose repeatedly between two alternatives and is rewarded by each correct choice. In general, each alternative yields the reward with fixed probability along the sequence of trials; for instance, the left alternative yields the reward with probability 0.7 and the right alternative, with probability 0.3. It was observed that, in this kind of task, humans rarely achieve optimal performance. Besides, subjects belonging to different age bands employ different strategies and achieve different levels of performance, both when each alternative yields the reward with fixed probability and when the sequence of correct choices is generated by Markov chains (VICTORINO, 2012). Such results may be due to the decline in the activity of the dopaminergic system and the basal ganglia along a lifetime, given that, in decision making, these structures are related to the expectation of reward and the evaluation of results. The present project is justified by the need to understand better the neurophysiological mechanisms that underlie human decision making, a particularly relevant aspect being the ontogenetic establishment of decisional strategies. To such an end, we propose to carry out computational simulations and decisional experiments in human subjects. The computational experiments consist in modeling the relevant aspects of the basal ganglia and decisional ontogenesis by means of artificial neural networks in order to reproduce the conditions associated to different age bands in repeated binary choice tasks. The results obtained by the use of this model will help us to understand the results obtained in experiments with humans. The decisional experiments in human subjects seek to evaluate the influence of different reward and punishment regimens in decision making. The task will also be performed in such a way to make clear the involved probabilities and the role of randomness in the determination of the correct choice. Thus, we seek to decrease the behavior of searching for patterns in the sequence of correct choices, widely reported in previous studies.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

DA SILVA, CAROLINA FEHER; SILVA MORGERO, KELLY CRISTINA; MOTA, AMANDA MANZINI; PIMENTEL PIEMONTE, MARIA ELISA; CHRYSOSTOMO BALDO, MARCUS VINICIUS. Aging and Parkinson's disease as functional models of temporal order perception. Neuropsychologia, v. 78, p. 1-9, NOV 2015. (03/11139-8, 13/10694-0)

DA SILVA, CAROLINA FEHER; VICTORINO, CAMILA GOMES; CATICHA, NESTOR; BALDO, MARCUS VINICIUS CHRYSOSTOMO. Exploration and recency as the main proximate causes of probability matching: a reinforcement learning analysis. SCIENTIFIC REPORTS, v. 7, NOV 10 2017. (13/10694-0, 13/13352-2)

DA SILVA, CAROLINA FEHER; BALDO, MARCUS V. C.. Computational models of the Posner simple and choice reaction time tasks. FRONTIERS IN COMPUTATIONAL NEUROSCIENCE, v. 9, JUL 2 2015. (13/13352-2, 13/10694-0)