Thomas Boraud - Academia.edu (original) (raw)
Papers by Thomas Boraud
Experimental Brain Research, 2004
The kinematics of straight reaching movements can be specified vectorially by the direction of th... more The kinematics of straight reaching movements can be specified vectorially by the direction of the movement and its extent. To explore the representation in the brain of these two properties, psychophysical studies have examined learning of visuomotor transformations of either rotation or gain and their generalization. However, the neuronal substrates of such complex learning are only beginning to be addressed. As an initial step in ensuring the validity of such investigations, it must be shown that monkeys indeed learn and generalize visuomotor transformations in the same manner as humans. Here, we analyze trajectories and velocities of movements as monkeys adapt to either rotational or gain transformations. We used rotations with different signs and magnitudes, and gains with different signs, and analyzed transfer of learning to untrained movements. The results show that monkeys can adapt to both types of transformation with a time course that resembles human learning. Analysis of the aftereffects reveals that rotation is learned locally and generalizes poorly to untrained directions, whereas gain is learned more globally and can be transferred to other amplitudes. The results lend additional support to the hypothesis that reaching movements are learned locally but can be easily rescaled to other magnitudes by scaling the peak velocity. The findings also indicate that reaching movements in monkeys are planned and executed very similarly to those in humans. This validates the underlying presumption that neuronal recordings in primates can help elucidate the mechanisms of motor learning in particular and motor planning in general.
Action-outcome (A-O) and stimulus-response (S-R) processes that are two forms of instrumental con... more Action-outcome (A-O) and stimulus-response (S-R) processes that are two forms of instrumental conditioning that are important components of decision making and action selection. The former adapts its response according to the outcome while the latter is insensitive to the outcome. An unsolved question is how these two processes emerge, cooperate and interact inside the brain in order to issue a unique behavioral answer. Here we propose a model of the interaction between the cortex, the basal ganglia and the thalamus based on a dual competition. We hypothesize that the striatum, the subthalamic nucleus, the internal pallidum (GPi), the thalamus, and the cortex are involved in closed feedback loops through the hyperdirect and direct pathways. These loops support a competition process that results in the ability for the basal ganglia to make a cognitive decision followed by a motor decision. Considering lateral cortical interactions (short range excitation, long range inhibition), anot...
Primary motor cortex (MI) neurons discharge vigorously during voluntary movement. A cardinal symp... more Primary motor cortex (MI) neurons discharge vigorously during voluntary movement. A cardinal symptom of Parkinson's disease (PD) is poverty of movement (akinesia). Current models of PD thus hypothesize that increased inhibitory pallidal output reduces firing rates in frontal cortex, including MI, resulting in akinesia and muscle rigidity. We recorded the simultaneous spontaneous discharge of several neurons in the arm-related area of MI of two monkeys and in the globus pallidus (GP) of one of the two. Accelerometers were fastened to the forelimbs to detect movement, and surface electromyograms were recorded from the contralateral arm of one monkey. The recordings were conducted before and after systemic treatment with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), rendering the animals severely akinetic and rigid with little or no tremor. The mean spontaneous MI rates during periods of immobility (four to five spikes/sec) did not change after MPTP; however, in this parkinsonian state, MI neurons discharged in long bursts (sometimes Ͼ2 sec long). These bursts were synchronized across many cells but failed to elicit detectable movement, indicating that even robust synchronous MI discharge need not result in movement. These synchronized population bursts were absent from the GP and were on a larger timescale than oscillatory synchrony found in the GP of tremulous MPTP primates, suggesting that MI parkinsonian synchrony arises independently of basal ganglia dynamics. After MPTP, MI neurons responded more vigorously and with less specificity to passive limb movement. Abnormal MI firing patterns and synchronization, rather than reduced firing rates, may underlie PD akinesia and persistent muscle rigidity.
It is now widely believed that decisions are guided by a small number of internal subjective vari... more It is now widely believed that decisions are guided by a small number of internal subjective variables that determine choice preference. The process of learning manifests as a change in the state of these variables. It is not clear how to find the neural correlates of these variables, in particular because their state cannot be directly measured or controlled by the experimenter. Rather, these variables reflect the history of the subject’s actions and reward experience. We seek to construct a behavioral model that captures the dynamics of learning and decision making, such that the internal variables of this model will serve as a proxy for the subjective variables. We use the theory of reinforcement learning in order to find a behavioral model that best captures the learning dynamics of monkeys in a two-armed bandit reward schedule. We consider two families of learning algorithms: value function estimation and direct policy optimization. In the former, the values of the alternative ...
Both computer science and behavioural psychology have proposed general theories to explain decisi... more Both computer science and behavioural psychology have proposed general theories to explain decision making behaviour in a stochastic environment. These theories are based on experiments in which a subject repeatedly chooses from a set of alternatives and is repeatedly rewarded according to his choices. These studies commonly come to the conclusion that decision learning behaviour is modulated by the reward framework properties. But the precise mechanisms underlying these behavioural modulations are still debated. Aiming to bridge between psychological descriptions of decision making behaviour and the potential computational mechanisms underlying this behaviour, we investigated this question and we present results based on data collected from a task in which 2 monkeys had to learn new associations of reward probabilities and targets randomly presented in 4 directions in a two choices decision task.
Movement disorders : official journal of the Movement Disorder Society, Jan 22, 2016
There is an apparent contradiction between experimental data showing that the basal ganglia are i... more There is an apparent contradiction between experimental data showing that the basal ganglia are involved in goal-oriented and routine behaviors and clinical observations. Lesion or disruption by deep brain stimulation of the globus pallidus interna has been used for various therapeutic purposes ranging from the improvement of dystonia to the treatment of Tourette's syndrome. None of these approaches has reported any severe impairment in goal-oriented or automatic movement. To solve this conundrum, we trained 2 monkeys to perform a variant of a 2-armed bandit-task (with different reward contingencies). In the latter we alternated blocks of trials with choices between familiar rewarded targets that elicit routine behavior and blocks with novel pairs of targets that require an intentional learning process. Bilateral inactivation of the globus pallidus interna, by injection of muscimol, prevents animals from learning new contingencies while performance remains intact, although slowe...
European Journal of Neuroscience, Jun 1, 1999
The purpose of the present study was to investigate whether the subthalamic nucleus (STN) was imp... more The purpose of the present study was to investigate whether the subthalamic nucleus (STN) was implicated in the glutamatergic compensatory mechanisms which have been shown to mask the parkinsonian motor abnormalities at the end of the presymptomatic period in experimental parkinsonism. Using multiunit electrophysiological recordings, we follow changes of activity occurring in the STN and in both the pars externalis and the pars internalis of the globus pallidus of monkeys chronically intoxicated with 1-methyl-4phenyl-1,2,3,6-tetrahydropyridine (MPTP), according to a protocol designed to mimic the gradual evolution of dopaminergic neuronal death. STN activity augmented signi®cantly in the course of treatment, even before the ®rst appearance of clinical signs (P < 0.01). This result would indicate that the STN, which increases its level of activity even before the end of the presymptomatic period, is principally responsible for the instigation of glutamatergic compensatory mechanisms which allow the maintenance of the striatal dopaminergic homeostasis.
Progress in Neurobiology, Dec 31, 1999
AbstractÐParkinson's disease remains one of the greatest challenges facing those who work in the ... more AbstractÐParkinson's disease remains one of the greatest challenges facing those who work in the ®eld of neurological research. Although the development of levodopa treatment revolutionised management of this debilitating diseases, no eective long-term therapy has yet been found.
Journal of Neurochemistry, Feb 1, 2011
D 2 -like antagonists potentiate dopamine release. They also inhibit dopamine uptake by a mechani... more D 2 -like antagonists potentiate dopamine release. They also inhibit dopamine uptake by a mechanism yet to be clarified. Here, we monitored dopamine uptake in the striatum of anesthetized mice. The dopamine overflow was evoked by brief electrical stimulation of the medial forebrain bundle (four pulses at 100 Hz) and was monitored with carbon fiber electrodes combined with continuous amperometry. The decay phase of evoked overflows reflects dopamine half-life, which entirely depends on uptake. The D 2 -like antagonists haloperidol and eticlopride enhanced the half-life by 45% and 48%, respectively, a moderate effect as compared to the uptake blocker nomifensine (528%). Both D 2 -like antagonists did not affect dopamine uptake in mice lacking D 2 receptors. Inhibi-tion of tonic dopamine release by gamma-butyrolactone did not mimic the enhancing effect of D 2 antagonists on dopamine half-life. However, prolonged stimulation boosted dopamine uptake and this effect was not observed after haloperidol treatment or in mice lacking D 2 receptors. Therefore, dopamine uptake is accelerated in conditions of excessive D 2 stimulation but not finely tuned in resting conditions. Inhibition of dopamine uptake by D 2 antagonists synergizes with the potentiation of dopamine release to strongly alter the phasic dopamine signaling.
Brain Res, 1998
We studied the effects of l-DOPA on the firing patterns of pallidal neurons in experimental parki... more We studied the effects of l-DOPA on the firing patterns of pallidal neurons in experimental parkinsonism. After a unilateral injection of MPTP, we observed a decrease in the firing rate of GPe neurons, and a slight increase in their bursting activity. In the GPi, there was a considerable augmentation of both neuronal firing frequency and the number of bursting cells. During l-DOPA treatment (10 mg/kg), GPe neurons.pattern is almost unmodified. The firing frequency of GPi neurons, on the contrary, decreased even lower than the control level. A slight reduction was observed in bursting activity. These unexpected results would show that the normalizing effect of l-DOPA on GPi output is limited.
The decision making process is a phenomenon which is hard to track using the standard signal aver... more The decision making process is a phenomenon which is hard to track using the standard signal averaging methods such as peri-event time histograms (PETHs). Indeed, even if the sequence of the events is controlled during a behavioural task, the inter-event interval duration remains highly variable. We have applied a temporal normalization method such that PETHs can be mapped across all events of a task trial and compared from neuron to neuron and from session to session. We have also applied Shannon mutual information theory to compare neural activity in recorded striatum and GPi neurons in behaving monkeys during a centre-out motor task and investigated individual neuron coding properties. These methods illustrate information pathways and their relative involvement. We also show that the GPi neurons recorded during the decision phase contained more information about the context and choice combination than striatal neurons.
Neural recording and processing systems are keyinstruments for investigating brain mechanisms and... more Neural recording and processing systems are keyinstruments for investigating brain mechanisms and exploring innovative solutions for neuroprosthetic devices. At early experimental stages, configurability is an important feature for the neural processing units. We present in this paper the design of a configurable ASIC, dedicated to the analysis of Local Field Potential spectral distribution in rats' deep brain region. We detail the filter bank design strategy, regarding specifications such as flexibility, low power consumption and high linearity range. Simulation results are presented with numerical and in vivo recorded neural signals.
Decision making takes a crucial role in daily life and can be seriously impaired in neurodegenera... more Decision making takes a crucial role in daily life and can be seriously impaired in neurodegenerative diseases. It implies cortical-basal ganglia loops, but the respective roles of each structure are still debated. Urodeles offer a remarkable opportunity to address this question. First, whereas the organization of the basal ganglia in these animals is similar to the one of mammals, it has orders of magnitudes fewer neurons1 and is therefore at a level of complexity that is more tractable in terms of comprehension and modelling. Second, the extensive regenerative capacities of their midbrain networks allow the use of local lesions to investigate their connectivity and functionality. This also makes the urodele a unique experimental model for investigating the mechanisms of decision-making network plasticity after a midbrain injury. Therefore, we decided to set up a behavioral test of decision making in this animal model. Our task was based on the ability of urodele to discriminate po...
If basal ganglia are widely accepted to participate in the high-level cognitive function of decis... more If basal ganglia are widely accepted to participate in the high-level cognitive function of decision-making, their role is less clear regarding the formation of habits. One of the biggest problem is to understand how goal-directed actions are transformed into habitual responses, or, said differently, how an animal can shift from an action-outcome (A-O) system to a stimulus-response (S-R) one while keeping a consistent behaviour. We introduce a computational model (basal ganglia, thalamus and cortex) that can solve a simple two arm-bandit task using reinforcement learning and explicit valuation of the outcome (Guthrie et al. (2013)). Hebbian learning has been added at the cortical level such that the model learns each time a move is issued, rewarded or not. Then, by inhibiting the output nuclei of the model (GPi), we show how learning has been transferred from the basal ganglia to the cortex, simply as a consequence of the statistics of the choice. Because best (in the sense of most ...
Et si notre faculté à prendre des décisions relevait plus du hasard que d’un processus rationnel ... more Et si notre faculté à prendre des décisions relevait plus du hasard que d’un processus rationnel ? On a longtemps admis que, chez l’homme, la prise de décision résultait d’un processus cognitif et psychologique : l’esprit décide, le corps obéit. Or, le schéma est inverse : le mécanisme décisionnel est produit par la matière cérébrale. C’est un phénomène aléatoire qui résulte de processus de compétitions au sein d’un réseau dont l’architecture a peu évolué depuis les premiers vertébrés. L’extraordinaire développement du cortex, qui a rendu possible le développement de grandes capacités d’abstraction, n’a pas modifié la structure initiale du réseau de la décision : le processus conserve sa nature aléatoire, ce qui limite la capacité de l’homo sapiens à raisonner de façon rationnelle. Il en résulte que lorsqu’un individu pèse le pour et le contre, il ne fait ni plus ni moins que de s’en remettre au hasard de dés virtuels. Apprendre consiste dès lors à piper ces dés en sa faveur… Mais c...
It is now widely believed that decisions are guided by a small number of internal subjective vari... more It is now widely believed that decisions are guided by a small number of internal subjective variables that determine choice preference. The process of learning manifests as a change in the state of these variables. It is not clear how to find the neural correlates of these variables, in particular because their state cannot be directly measured or controlled by the experimenter. Rather, these variables reflect the history of the subject’s actions and reward experience. We seek to construct a behavioral model that captures the dynamics of learning and decision making, such that the internal variables of this model will serve as a proxy for the subjective variables. We use the theory of reinforcement learning in order to find a behavioral model that best captures the learning dynamics of monkeys in a two-armed bandit reward schedule. We consider two families of learning algorithms: value function estimation and direct policy optimization. In the former, the values of the alternative ...
Experimental Brain Research, 2004
The kinematics of straight reaching movements can be specified vectorially by the direction of th... more The kinematics of straight reaching movements can be specified vectorially by the direction of the movement and its extent. To explore the representation in the brain of these two properties, psychophysical studies have examined learning of visuomotor transformations of either rotation or gain and their generalization. However, the neuronal substrates of such complex learning are only beginning to be addressed. As an initial step in ensuring the validity of such investigations, it must be shown that monkeys indeed learn and generalize visuomotor transformations in the same manner as humans. Here, we analyze trajectories and velocities of movements as monkeys adapt to either rotational or gain transformations. We used rotations with different signs and magnitudes, and gains with different signs, and analyzed transfer of learning to untrained movements. The results show that monkeys can adapt to both types of transformation with a time course that resembles human learning. Analysis of the aftereffects reveals that rotation is learned locally and generalizes poorly to untrained directions, whereas gain is learned more globally and can be transferred to other amplitudes. The results lend additional support to the hypothesis that reaching movements are learned locally but can be easily rescaled to other magnitudes by scaling the peak velocity. The findings also indicate that reaching movements in monkeys are planned and executed very similarly to those in humans. This validates the underlying presumption that neuronal recordings in primates can help elucidate the mechanisms of motor learning in particular and motor planning in general.
Action-outcome (A-O) and stimulus-response (S-R) processes that are two forms of instrumental con... more Action-outcome (A-O) and stimulus-response (S-R) processes that are two forms of instrumental conditioning that are important components of decision making and action selection. The former adapts its response according to the outcome while the latter is insensitive to the outcome. An unsolved question is how these two processes emerge, cooperate and interact inside the brain in order to issue a unique behavioral answer. Here we propose a model of the interaction between the cortex, the basal ganglia and the thalamus based on a dual competition. We hypothesize that the striatum, the subthalamic nucleus, the internal pallidum (GPi), the thalamus, and the cortex are involved in closed feedback loops through the hyperdirect and direct pathways. These loops support a competition process that results in the ability for the basal ganglia to make a cognitive decision followed by a motor decision. Considering lateral cortical interactions (short range excitation, long range inhibition), anot...
Primary motor cortex (MI) neurons discharge vigorously during voluntary movement. A cardinal symp... more Primary motor cortex (MI) neurons discharge vigorously during voluntary movement. A cardinal symptom of Parkinson's disease (PD) is poverty of movement (akinesia). Current models of PD thus hypothesize that increased inhibitory pallidal output reduces firing rates in frontal cortex, including MI, resulting in akinesia and muscle rigidity. We recorded the simultaneous spontaneous discharge of several neurons in the arm-related area of MI of two monkeys and in the globus pallidus (GP) of one of the two. Accelerometers were fastened to the forelimbs to detect movement, and surface electromyograms were recorded from the contralateral arm of one monkey. The recordings were conducted before and after systemic treatment with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), rendering the animals severely akinetic and rigid with little or no tremor. The mean spontaneous MI rates during periods of immobility (four to five spikes/sec) did not change after MPTP; however, in this parkinsonian state, MI neurons discharged in long bursts (sometimes Ͼ2 sec long). These bursts were synchronized across many cells but failed to elicit detectable movement, indicating that even robust synchronous MI discharge need not result in movement. These synchronized population bursts were absent from the GP and were on a larger timescale than oscillatory synchrony found in the GP of tremulous MPTP primates, suggesting that MI parkinsonian synchrony arises independently of basal ganglia dynamics. After MPTP, MI neurons responded more vigorously and with less specificity to passive limb movement. Abnormal MI firing patterns and synchronization, rather than reduced firing rates, may underlie PD akinesia and persistent muscle rigidity.
It is now widely believed that decisions are guided by a small number of internal subjective vari... more It is now widely believed that decisions are guided by a small number of internal subjective variables that determine choice preference. The process of learning manifests as a change in the state of these variables. It is not clear how to find the neural correlates of these variables, in particular because their state cannot be directly measured or controlled by the experimenter. Rather, these variables reflect the history of the subject’s actions and reward experience. We seek to construct a behavioral model that captures the dynamics of learning and decision making, such that the internal variables of this model will serve as a proxy for the subjective variables. We use the theory of reinforcement learning in order to find a behavioral model that best captures the learning dynamics of monkeys in a two-armed bandit reward schedule. We consider two families of learning algorithms: value function estimation and direct policy optimization. In the former, the values of the alternative ...
Both computer science and behavioural psychology have proposed general theories to explain decisi... more Both computer science and behavioural psychology have proposed general theories to explain decision making behaviour in a stochastic environment. These theories are based on experiments in which a subject repeatedly chooses from a set of alternatives and is repeatedly rewarded according to his choices. These studies commonly come to the conclusion that decision learning behaviour is modulated by the reward framework properties. But the precise mechanisms underlying these behavioural modulations are still debated. Aiming to bridge between psychological descriptions of decision making behaviour and the potential computational mechanisms underlying this behaviour, we investigated this question and we present results based on data collected from a task in which 2 monkeys had to learn new associations of reward probabilities and targets randomly presented in 4 directions in a two choices decision task.
Movement disorders : official journal of the Movement Disorder Society, Jan 22, 2016
There is an apparent contradiction between experimental data showing that the basal ganglia are i... more There is an apparent contradiction between experimental data showing that the basal ganglia are involved in goal-oriented and routine behaviors and clinical observations. Lesion or disruption by deep brain stimulation of the globus pallidus interna has been used for various therapeutic purposes ranging from the improvement of dystonia to the treatment of Tourette's syndrome. None of these approaches has reported any severe impairment in goal-oriented or automatic movement. To solve this conundrum, we trained 2 monkeys to perform a variant of a 2-armed bandit-task (with different reward contingencies). In the latter we alternated blocks of trials with choices between familiar rewarded targets that elicit routine behavior and blocks with novel pairs of targets that require an intentional learning process. Bilateral inactivation of the globus pallidus interna, by injection of muscimol, prevents animals from learning new contingencies while performance remains intact, although slowe...
European Journal of Neuroscience, Jun 1, 1999
The purpose of the present study was to investigate whether the subthalamic nucleus (STN) was imp... more The purpose of the present study was to investigate whether the subthalamic nucleus (STN) was implicated in the glutamatergic compensatory mechanisms which have been shown to mask the parkinsonian motor abnormalities at the end of the presymptomatic period in experimental parkinsonism. Using multiunit electrophysiological recordings, we follow changes of activity occurring in the STN and in both the pars externalis and the pars internalis of the globus pallidus of monkeys chronically intoxicated with 1-methyl-4phenyl-1,2,3,6-tetrahydropyridine (MPTP), according to a protocol designed to mimic the gradual evolution of dopaminergic neuronal death. STN activity augmented signi®cantly in the course of treatment, even before the ®rst appearance of clinical signs (P < 0.01). This result would indicate that the STN, which increases its level of activity even before the end of the presymptomatic period, is principally responsible for the instigation of glutamatergic compensatory mechanisms which allow the maintenance of the striatal dopaminergic homeostasis.
Progress in Neurobiology, Dec 31, 1999
AbstractÐParkinson's disease remains one of the greatest challenges facing those who work in the ... more AbstractÐParkinson's disease remains one of the greatest challenges facing those who work in the ®eld of neurological research. Although the development of levodopa treatment revolutionised management of this debilitating diseases, no eective long-term therapy has yet been found.
Journal of Neurochemistry, Feb 1, 2011
D 2 -like antagonists potentiate dopamine release. They also inhibit dopamine uptake by a mechani... more D 2 -like antagonists potentiate dopamine release. They also inhibit dopamine uptake by a mechanism yet to be clarified. Here, we monitored dopamine uptake in the striatum of anesthetized mice. The dopamine overflow was evoked by brief electrical stimulation of the medial forebrain bundle (four pulses at 100 Hz) and was monitored with carbon fiber electrodes combined with continuous amperometry. The decay phase of evoked overflows reflects dopamine half-life, which entirely depends on uptake. The D 2 -like antagonists haloperidol and eticlopride enhanced the half-life by 45% and 48%, respectively, a moderate effect as compared to the uptake blocker nomifensine (528%). Both D 2 -like antagonists did not affect dopamine uptake in mice lacking D 2 receptors. Inhibi-tion of tonic dopamine release by gamma-butyrolactone did not mimic the enhancing effect of D 2 antagonists on dopamine half-life. However, prolonged stimulation boosted dopamine uptake and this effect was not observed after haloperidol treatment or in mice lacking D 2 receptors. Therefore, dopamine uptake is accelerated in conditions of excessive D 2 stimulation but not finely tuned in resting conditions. Inhibition of dopamine uptake by D 2 antagonists synergizes with the potentiation of dopamine release to strongly alter the phasic dopamine signaling.
Brain Res, 1998
We studied the effects of l-DOPA on the firing patterns of pallidal neurons in experimental parki... more We studied the effects of l-DOPA on the firing patterns of pallidal neurons in experimental parkinsonism. After a unilateral injection of MPTP, we observed a decrease in the firing rate of GPe neurons, and a slight increase in their bursting activity. In the GPi, there was a considerable augmentation of both neuronal firing frequency and the number of bursting cells. During l-DOPA treatment (10 mg/kg), GPe neurons.pattern is almost unmodified. The firing frequency of GPi neurons, on the contrary, decreased even lower than the control level. A slight reduction was observed in bursting activity. These unexpected results would show that the normalizing effect of l-DOPA on GPi output is limited.
The decision making process is a phenomenon which is hard to track using the standard signal aver... more The decision making process is a phenomenon which is hard to track using the standard signal averaging methods such as peri-event time histograms (PETHs). Indeed, even if the sequence of the events is controlled during a behavioural task, the inter-event interval duration remains highly variable. We have applied a temporal normalization method such that PETHs can be mapped across all events of a task trial and compared from neuron to neuron and from session to session. We have also applied Shannon mutual information theory to compare neural activity in recorded striatum and GPi neurons in behaving monkeys during a centre-out motor task and investigated individual neuron coding properties. These methods illustrate information pathways and their relative involvement. We also show that the GPi neurons recorded during the decision phase contained more information about the context and choice combination than striatal neurons.
Neural recording and processing systems are keyinstruments for investigating brain mechanisms and... more Neural recording and processing systems are keyinstruments for investigating brain mechanisms and exploring innovative solutions for neuroprosthetic devices. At early experimental stages, configurability is an important feature for the neural processing units. We present in this paper the design of a configurable ASIC, dedicated to the analysis of Local Field Potential spectral distribution in rats' deep brain region. We detail the filter bank design strategy, regarding specifications such as flexibility, low power consumption and high linearity range. Simulation results are presented with numerical and in vivo recorded neural signals.
Decision making takes a crucial role in daily life and can be seriously impaired in neurodegenera... more Decision making takes a crucial role in daily life and can be seriously impaired in neurodegenerative diseases. It implies cortical-basal ganglia loops, but the respective roles of each structure are still debated. Urodeles offer a remarkable opportunity to address this question. First, whereas the organization of the basal ganglia in these animals is similar to the one of mammals, it has orders of magnitudes fewer neurons1 and is therefore at a level of complexity that is more tractable in terms of comprehension and modelling. Second, the extensive regenerative capacities of their midbrain networks allow the use of local lesions to investigate their connectivity and functionality. This also makes the urodele a unique experimental model for investigating the mechanisms of decision-making network plasticity after a midbrain injury. Therefore, we decided to set up a behavioral test of decision making in this animal model. Our task was based on the ability of urodele to discriminate po...
If basal ganglia are widely accepted to participate in the high-level cognitive function of decis... more If basal ganglia are widely accepted to participate in the high-level cognitive function of decision-making, their role is less clear regarding the formation of habits. One of the biggest problem is to understand how goal-directed actions are transformed into habitual responses, or, said differently, how an animal can shift from an action-outcome (A-O) system to a stimulus-response (S-R) one while keeping a consistent behaviour. We introduce a computational model (basal ganglia, thalamus and cortex) that can solve a simple two arm-bandit task using reinforcement learning and explicit valuation of the outcome (Guthrie et al. (2013)). Hebbian learning has been added at the cortical level such that the model learns each time a move is issued, rewarded or not. Then, by inhibiting the output nuclei of the model (GPi), we show how learning has been transferred from the basal ganglia to the cortex, simply as a consequence of the statistics of the choice. Because best (in the sense of most ...
Et si notre faculté à prendre des décisions relevait plus du hasard que d’un processus rationnel ... more Et si notre faculté à prendre des décisions relevait plus du hasard que d’un processus rationnel ? On a longtemps admis que, chez l’homme, la prise de décision résultait d’un processus cognitif et psychologique : l’esprit décide, le corps obéit. Or, le schéma est inverse : le mécanisme décisionnel est produit par la matière cérébrale. C’est un phénomène aléatoire qui résulte de processus de compétitions au sein d’un réseau dont l’architecture a peu évolué depuis les premiers vertébrés. L’extraordinaire développement du cortex, qui a rendu possible le développement de grandes capacités d’abstraction, n’a pas modifié la structure initiale du réseau de la décision : le processus conserve sa nature aléatoire, ce qui limite la capacité de l’homo sapiens à raisonner de façon rationnelle. Il en résulte que lorsqu’un individu pèse le pour et le contre, il ne fait ni plus ni moins que de s’en remettre au hasard de dés virtuels. Apprendre consiste dès lors à piper ces dés en sa faveur… Mais c...
It is now widely believed that decisions are guided by a small number of internal subjective vari... more It is now widely believed that decisions are guided by a small number of internal subjective variables that determine choice preference. The process of learning manifests as a change in the state of these variables. It is not clear how to find the neural correlates of these variables, in particular because their state cannot be directly measured or controlled by the experimenter. Rather, these variables reflect the history of the subject’s actions and reward experience. We seek to construct a behavioral model that captures the dynamics of learning and decision making, such that the internal variables of this model will serve as a proxy for the subjective variables. We use the theory of reinforcement learning in order to find a behavioral model that best captures the learning dynamics of monkeys in a two-armed bandit reward schedule. We consider two families of learning algorithms: value function estimation and direct policy optimization. In the former, the values of the alternative ...