doi:10.1016/S0925-2312(02)00813-5 
Copyright © 2003 Published by Elsevier Science B.V.
The role of cortico-basal-thalamic loops in cognition: a computational model and preliminary results
ICSI, 1947 Center Street, Berkeley 94704, USA
Available online 28 March 2003.
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Abstract
Clinical and experimental research over the last decade has implicated neuroanatomic loops connecting the frontal cortex to the basal ganglia and thalamus in various aspects of planning and memory. We report on computational model whose central aspects are: (1) a model of cortical-striatal-thalamic loops in planning and executive control, and (2) a fine-grained model of basal-ganglia function that exploits specific component connectivity and dynamics. The model is biologically plausible given current literature on the neurophysiology and disease pathology of the relevant brain regions. Specifically, our model has implications for subjects with diseases affecting the relevant brain regions (Parkinson's disease and Huntington's disease).
Fig. 1. This shows the block diagram view of the components and connectivity of the Cortico-basal loops. Inhibitory projections are shown with rounded tips while excitatory connections are shown with pointy tipped arrows. The direct and indirect pathways through the BG complex are highlighted and identified within the figure. The Striatum is shown with the two types of dopamine receptors D1 and D2. The Striatum, STN, GP (GPe and GPi), STN and SN (SNPr and SNPc) together comprise the primate BG complex.
Fig. 2. The Basic Model of collections of BG cells. The input sites may be excitatory, inhibitory or weighted (the resource link). The firing of cells can be conjunctive (there is excitation in all the excitatory cites, no inhibition and sufficient activation in the resource arcs, then the cell fires). Weights on the output links (default 1) represent the
Gain parameter. Different regions in the BG are modeled by different types of nodes in the model. The striatal cells in the model are conjunctive matching cells. Other excitatory nuclei (such as STN) use the logistic firing function used in many neural network models. Other inhibitory cells (such as the output nuclei) use other firing functions based on the exponential family of functions. Also, different deficits of specific regions in the BG are modeled changing the gain, firing function and weights of the relevant region in the model.
Fig. 3. Implementation and computational simulation of the BG complex. The figure shows part of the Basal Ganglia model. The striatal match cells are implemented as conjunctive transitions (shown as CorticalStriatal (CS) transitions in the Stochastic Petri Net based model). The striatal cells are bi-stable, and when in the up state can inhibit the output nuclei (the St-Gpi transitions) with input. This is the direct pathway which results in disinhibiting the thalamus (through the BG-thal transitions in the model. The inhibitory effect of the indirect pathway can be seen through the GPe cells. Other pathways include the direct stimulation of STN and the STN-GPe loops.
Fig. 4. The model's performance of the different patient data. The PERS label in the chart shows the relative performance of different populations in terms of the perseverative errors. Notice that the model predicts high percentage of perseverative errors for the HD and Schizophrenic patients compared to the normals. Notice also that the normals perform quite well in that the get all the categories right (percentage of categories right is shown by the value corresponding to the label CAT. Notice as well that the model performs badly for all other patient populations.
Fig. 5. (a) Performance of Model (M) compared to published data (from [
1]) on the
PD patient population. The chart shows the comparison of the model predictions with the data for patients with and without dementia. (b) Performance of Model (M) compared to published data (from [
1]) on the
HD patient population. The chart shows the comparison of the model predictions with the data for patients with and without dementia.
Abstract
Purchase PDF (427 K)
E-mail Article
Add to my Quick Links
Cited By in Scopus (0)
Purchase PDF (5313 K)
