tract anatomy

index.md

@@ -76,7 +76,7 @@ The thalamocortical structural connectivity tractography atlas was generated in
 * **Delineate regionally-specific thalamus-to-cortex connections**: Structural connections between the thalamus and ipsilateral cortical regions were extracted from the thalamic tractography with [thalamocortical_structuralconnectivity/template/thalamocortical_connectoms.sh](https://github.com/PennLINC/thalamocortical_development/blob/main/thalamocortical_structuralconnectivity/template/thalamocortical_connections.sh) using the HCP-MMP (glasser) atlas.  
 * **Manually curate thalamocortical connections**: All regionally-specific thalamocortical connections were visualized and manually edited, if needed, to ensure that atlas connections were compact, robust, comparable across hemispheres, and anatomically correct (based on available primate tract tracing or human diffusion MRI data). Manual editing included the removal of clear false positive streamlines and facilitated identification of thalamocortical connections that were very sparse or unreliable for removal from the atlas.
 * **Create skeletonized connections for autotracking**: In order to facilitate use of the thalamocortical structural connectivity atlas with dsi-studio's autotrack, each regionally-specific connection was skeletonized by deleting repeat streamlines with [thalamocortical_structuralconnectivity/template/sparsify_connections.sh](https://github.com/PennLINC/thalamocortical_development/blob/main/thalamocortical_structuralconnectivity/template/sparsify_connections.sh).  
-* **Combine all regionally-specific thalamocortical connections into one tractography atlas**: After generating finalized, regionally-specific structural connections between the thalamus and individual cortical areas, all connections were combined into one .tt.gz file for use in this study and for public distribution. The final version of the atlas only includes connections that could be robustly and reliably delineated in both the high-resolution HCPYA diffusion template and in individual participant's data in the PNC (single-shell, b=1000) and HCPD (multi-shell, bs=1500,3000). This atlas is provided in [/thalamocortical_autotrack_template](https://github.com/PennLINC/thalamocortical_development/tree/main/thalamocortical_autotrack_template).   
+* **Combine all regionally-specific thalamocortical connections into one tractography atlas**: After generating finalized, regionally-specific structural connections between the thalamus and individual cortical areas, all connections were combined into one .tt.gz file for use in this study and for public distribution. The final version of the atlas only includes connections that could be robustly and reliably delineated in both the high-resolution HCPYA diffusion template and in individual participant's data in the PNC (single-shell, b=1000) and HCPD (multi-shell, bs=1500,3000). This atlas is provided in [/results/thalamocortical_autotrack_template](https://github.com/PennLINC/thalamocortical_development/tree/main/results/thalamocortical_autotrack_template). 
 
 > To use this atlas with dsi-studio's autotrack to generate thalamocortical connections in individual participant data, both the "ICBM152_adult.tt.gz" (autotrack tracts) and "ICBM152_adult.tt.gz.txt" (tract name list) files are required and must be located in the location expected by the dsi-studio software.   
 >
@@ -147,7 +147,11 @@ To facilitate analysis of participant-level thalamocortical connectivity data, d
 
 * **Dataset-specific demographics and diffusion dfs**: Analytic dataframes were created for PNC and HCPD that include demographics and thalamocortical structural connectivity metrics for the final study sample. In HCPD, final sample thalamocortical metrics were harmonized with comBat to mitigate site effects. In both datasets, thalamocortical connections with < 5 streamlines (primary analysis) were removed from final analytic dataframes at the participant-level; these connections were not included in further analyses. These steps (dataframe creation, harmonization, and streamline thresholding) were implemented with [/sample_construction/PNC/tractmeasures_dfs_PNC.R](https://github.com/PennLINC/thalamocortical_development/blob/main/sample_construction/PNC/tractmeasures_dfs_PNC.R) and [sample_construction/HCPD/tractmeasures_dfs_HCPD.R](https://github.com/PennLINC/thalamocortical_development/blob/main/sample_construction/HCPD/tractmeasures_dfs_HCPD.R). 
 * **Dataset-specific tract lists**: Only connections that could be reliably and robustly delineated at the individual level were studied in this work. To identify thalamocortical connections to study in PNC and HCPD, [tract_measures/tractlists/thalamocortical_tractlists.R](https://github.com/PennLINC/thalamocortical_development/blob/main/tract_measures/tractlists/thalamocortical_tractlists.R) was run. For the primary analysis we implemented a connection-level streamline count threshold of >= 5 (as noted above) and a dataset-level connection inclusion threshold of >= 90% of participants. In PNC and HCPD, 6% and 3% of thalamocortical connections present in the atlas were excluded from analysis, respectively, as they were sparsely reconstructed in > 10% of participants. 
-
+
+### Thalamocortical Connectivity: Atlas Characteristics and Anatomical Gradients
+Thalamocortical connection anatomy was examined at the atlas and the group level to understand characteristics of reconstructed thalamic connections and to uncover cortical and thalamic anatomical gradients. This exploration included quantifying cortical surface coverage for the thalamocortical atlas (atlas-based); comparing the surface area and sulcal depth of cortical parcels with versus without thalamic connections represented in the atlas (atlas-based); surveying variation in thalamocortical connectivity strength across Mesulam cortical types and the S-A axis (group-based in PNC and HCPD); and testing whether thalamic endpoint core-matrix gradients values differed across Mesulam types and the S-A axis (group-based in PNC and HCPD). These analyses were conducted in [/results/tract_anatomy/thalamocortical_connection_anatomy.Rmd](https://github.com/PennLINC/thalamocortical_development/blob/main/results/tract_anatomy/thalamocortical_connection_anatomy.Rmd); a knit version of thalamocortical_connection_anatomy.html can be viewed [here](https://htmlpreview.github.io/?https://github.com/PennLINC/thalamocortical_development/blob/main/results/tract_anatomy/thalamocortical_connection_anatomy.html). 
+* The total surface area and average sulcal depth of all glasser parcels was computed based on the fsaverage surface in [/results/tract_anatomy/parcel_anatomy/glasser_parcel_anatomy.sh](https://github.com/PennLINC/thalamocortical_development/blob/main/results/tract_anatomy/parcel_anatomy/glasser_parcel_anatomy.sh) and [/results/tract_anatomy/parcel_anatomy/glasser_parcel_anatomy.py](https://github.com/PennLINC/thalamocortical_development/blob/main/results/tract_anatomy/parcel_anatomy/glasser_parcel_anatomy.py). 
+
 
 ### Software Installation
 The following external software was used in this project: