The Michael Halassa Lab
Identifying the role of frontal thalamocortical interactions in decision making under uncertainty
The ability to resolve uncertainty is critical for optimizing choices and outcomes during decision making. While the prefrontal cortex is essential to convert incoming sensory inputs into cognitive control signals, real world sensory inputs are often associated with uncertainty (eg: noise). How the PFC handles such uncertainty currently remains unresolved. Recent studies in humans have shown that the engagement of the cognitive thalamus during decision making scales proportionately with the degree of uncertainty. Using a variety of circuit dissection and modelling approaches we are developing a mechanistic understanding of the thalamic signals and cell type specific microcircuits that regulate prefrontal activity in decision making under uncertainty. Such an endeavor is highly relevant for psychiatric disorders like Schizophrenia as structural and functional impairments of thalamocortical connectivity are known to be associated with impaired decision-making associated with this disorder. As such, thalamic modulation of PFC activity can be used as an access point for developing targeted therapeutics for Schizophrenia.
Identifying the role of thalamocortical interactions in economic decision making in dynamic environment
The ability to use different strategies or schemas to deal with dynamic environments in decision making is essential for optimizing choices and outcomes. Multiple cortical areas are essential for economic decision making. Value coding in the Orbital frontal cortex has been extensively studied. Prefrontal cortex is involved in flexible value coding and strategizing in decision making under dynamically changing environments. However, how those cortical-cortical and thalamocortical interactions are computing changing values, shifting strategies and executing decisions in such a challenging environment was unknown. To answer this question, we introduced parameterized decision-making tasks in mice in which the value of options are dynamically changing. Combined with circuit dissection and modelling approaches we are trying to understand how different cortical circuits and their interaction with thalamus are optimizing decisions in the dynamic environment.
While strategization and executive functions are an important component of impaired cognitive function in schizophrenia, thalamocortical connectivity is known to be associated with impaired cognitive functions this disorder. Leveraging genetic mice models carrying the identical genetic alterations in schizophrenia patients combined with the approaches described above, we are aiming to develop circuit based therapeutic methods for improving cognitive function in schizophrenia.
Explaining the computational advantage of adding a thalamic-like architecture to recurrent networks
We develop recurrent neural networks with a thalamus-like component and synaptic plasticity rules to model the thalamocortical interactions in cognitive flexibility. We find that the MD component is able to extract context information by integrating context-relevant traces over trials and to suppress context-irrelevant neurons in PFC. Incorporating the MD disjoints the contextual representations and enables efficient population coding in PFC, which shows the computational advantages in context switch and continual learning.
We are developing a series of reservoir networks that communicate through thalamic intermediaries to solve probabilistic learning tasks. We find a role for MD in enhancing cognitive flexibility by responding to individual cortices, and also in mediating cortical interactions through a trans-thalamic route. Predictions for the model are compared with human fMRI data from subjects performing the same tasks.
We further propose that thalamocortical-basal ganglia interaction serves as a system level solution for flexible and generalizable credit assignment. Specifically, we propose that basal ganglia guides the thalamocortical plasticity in two time scales to enable meta learning–the fast plasticity allows flexible contextual association while the slow plasticity develops a cortical representation that can generalize across context. Furthermore, our theory indicates an advantage of thalamocortical architectures in continual learning and working memory tasks.