Decision making mechanisms in a connectionist model of the basal ganglia

Martin Guthrie (1, 1) , Charlotte Héricé (2, 1, 1) , Radwa Khalil (1) , Maria Moftah (3) , Thomas Boraud (1, 1) , André Garenne (1, 1, 2)

1 IMN - Institut des Maladies Neurodégénératives [Bordeaux]
2 Mnemosyne - Mnemonic Synergy
3 Alexandria University [Alexandrie]

The mechanisms of decision making are generally thought to be under the control of a set of cortico-sub cortical loops. There are known to be several parallel functional loops through the basal ganglia connecting back to distinct areas of cortex, processing different modalities of decision making, including motor, cognitive and limbic. Due to convergence and divergence within the network though, these loops cannot be completely segregated. We use these properties to develop a connectionist model at a spiking neuron level on the bases of the recently published Guthrie's model [1]. This model is demonstrated on a decision making task that has been studied in primates and the electrophysiology interpreted to show that the decision is made in two successive steps. In this task, the animals are trained to associate recompense values to targets in order to maximize their rewards [2]. To develop this model, we use two parallel loops, each of which performs decision making based on interactions between positive and negative feedback pathways within the loop. The loops communicate via divergence in one specific area. This model is tested to perform two level decision making as in primates. The whole system is instantiated using leaky integrate-and-fire neurons and its architecture relies on commonly accepted data regarding the complex functional connectivity description between basal ganglia, cortex and thalamus.  We are currently developing the model capability to mimic the monkey behaviour and its learning skills. Our goal is to apply a bottom-up approach of the basal ganglia model development in which the learning of optimum decision making mechanisms will emerge from the closed-loop interaction between the neural circuitry and its sensory-motor interface. This will allow us (i) to avoid the arbitrary choice of a pre- existing machine-learning derivative model and also (ii) to have the possibility to investigate the cell-scale mechanisms impact on the whole model capacities. REFERENCES [1] Guthrie, M., Leblois, A., Garenne, A. & Boraud, T. Interaction between cognitive and motor cortico--‐ asal. Journal of neurophysiology 109, 3025-3040 (2013). [2] Pasquereau, B. et al. Shaping of motor responses by incentive values through the basal ganglia. The Journal of neuroscience: the official journal of the Society for Neuroscience 27, 1176‐1183, doi:10.1523/JNEUROSCI.3745‐06.2007 (2007).

Mots clés: connectionist model; action selection; decision making; modeling; Basal ganglia

Domaines: Neurosciences [q-bio.NC]; Modélisation et simulation; Réseau de neurones [cs.NE]

Martin Guthrie, Charlotte Héricé, Radwa Khalil, Maria Moftah, Thomas Boraud, et al.. Decision making mechanisms in a connectionist model of the basal ganglia. Fourth Symposium on Biology of Decision-Making (SBDM) 2014, May 2014, Paris, France. . ⟨hal-01102619⟩