Thalamocortical contributions to cognitive task activity
- Department of Psychological and Brain Sciences, University of Iowa, United States
- Cognitive Control Collaborative, University of Iowa, United States
- Iowa Neuroscience Institute, University of Iowa, United States
- Department of Psychiatry, University of Iowa, United States
- Brain and Mind Center, University of Sydney, Australia
- Center for Molecular and Behavioral Neuroscience, Rutgers University-Newark, United States
Figures
Figure 1 with 2 supplements
Low-dimensional organization of thalamic task-evoked response that supports multi-task performance.
(A) We decomposed the high-dimensional multi-task evoked activity matrix into low-dimensional spatial components in the human thalamus and a task-wide loading matrix. For a list of all tasks, see Supplementary file 1. (B) Spatial topography of the top 3 components from the multi-domain task battery (MDTB) dataset that explained 50% of the variance in the group averaged task activity matrix. For top components for the Nakai and Nishimoto (N&N) dataset, see Figure 1—figure supplement 1. For the loadings between tasks and components, see Figure 1—figure supplement 2. (C) Results from applying principal component analysis (PCA) to single subjects. For both the MDTB and N&N datasets, for individual subjects up to 57% of the variance across multiple tasks can be explained by the top 10 components. Error bars and shaded areas indicate standard error of the mean.
Figure 1—figure supplement 1
Spatial topography of the top three components from the Nakai and Nishimoto (N&N) dataset that explained about 50% of the variance in the group averaged task activity matrix.
Figure 1—figure supplement 2
Loadings between individual tasks and thalamic activity components.
(A) Loadings matrix for the top 3 activity components and individual tasks for both datasets. (B) Low-dimensional embedding of task-evoked responses, plotted by task loadings on each activity component. Each colored dot represents a single task.
Figure 2 with 2 supplements
Hub regions in the thalamus.
(A) Task hubs (CompW) and functional connectivity (FC) hubs (PC) in the thalamus. PC = participation coefficient. (B) Projecting task hub metrics onto the cortex via thalamocortical FC.
Figure 2—figure supplement 1
Control analyses for the task hub metric.
Spatial topography of the task hub metric calculated using the lower 10th to 20th activity components.
Figure 2—figure supplement 2
Conceptual explanation of the participation coefficient metric.
PC is a measure of the number of between-network connectivity connections (estimated from functional connectivity) for each thalamus voxel, normalized by their expected number of connections. If a voxel has connections uniformly distributed to all cortical networks, then its PC value will be close to 1; on the other hand, if connectivity is concentrated to a specific network, its PC value will be close to 0. For the example on the left, the black region only has connectivity concentrated with regions of interests (ROIs) in network B, resulting in a PC value of 0. For the example on the right, the connectivity is distributed to both network A and B, resulting in a higher PC value of 0.5.
Figure 3
Anatomical distribution of task hub estimates in the thalamus.
AN = anterior nucleus, VM = ventromedial, MD = mediodorsal, IL = intralaminar, VA = ventral anterior, VL = ventrolateral, VP = ventroposterior, PuM = medial pulvinar, LP = lateral posterior, LGN = lateral geniculate nucleus, MGN = medial geniculate nucleus; V=visual network, SM = somoatomotor, DA = dorsal attention, CO = cingulo-opercular, FP = frontoparietal, DF = default mode. Error bar indicates standard error of the mean.
Figure 4 with 1 supplement
Activity flow model predicts cortical task-evoked response patterns.
(A) Model for testing whether thalamic task-evoked activity can predict patterns of cortical task-evoked responses. (B) Unnormalized prediction accuracy. Cortical regions divided by seven cortical functional networks: Vis = visual, SM = somatomotor, DA = dorsal attention, CO = cingulo-opercular, DF = default mode, FP = frontal parietal; Null model: randomly permute the evoked pattern in the thalamus. (C) Noise normalized prediction accuracy. (D) Prediction accuracy normalized by the noise ceiling. (E) Unnormalized activity flow model prediction for 100 cortical regions of interests (ROIs). (E) Noise normalized activity flow model prediction for 100 cortical ROIs.
Figure 4—figure supplement 1
Noise ceiling of different brain regions and null models.
Error bar indicates standard error of the mean.
Figure 5
Simulating the thalamic lesion’s effect on activity flow model prediction.
(A) Artificial lesion of 20% of the thalamus voxels based on their percentile rank of task hub property: examination of the impact on cortical task-evoked activity prediction. (B) Subregions that showed greater reduction in prediction accuracy were primarily located in anterior, medial, and posterior thalamus. Shaded error indicates standard error of the mean.
Figure 6
Neuropsychological evaluations from 20 patients with focal thalamic lesions.
(A) Twelve patients exhibited multiple-domain (MM) impairment (negative z-scores) across multiple neuropsychological assessments. Eight patients exhibited no multi-domain (SM) impairment. (B) Lesion sites from patients with and without multi-domain impairment. (C) Mean and standard error of the mean of the reduction in activity flow model prediction after virtual lesions, plotted separately for virtual lesion sites that overlapped with MM or SM lesions. Error bar indicates standard error of the mean. Panel A reproduced from Figure 2B from Hwang et al., 2021, with permission.
Author response image 1
Author response image 2
Additional files
-
Supplementary file 1
List of task conditions.
- https://cdn.elifesciences.org/articles/81282/elife-81282-supp1-v2.zip
-
MDAR checklist
- https://cdn.elifesciences.org/articles/81282/elife-81282-mdarchecklist1-v2.docx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Thalamocortical contributions to cognitive task activity
eLife 11:e81282.
https://doi.org/10.7554/eLife.81282
