The Human Connectome Project 7T Retinotopy Database
This page serves as a supplement to the 2018 VSS Poster and SfN Presentation on the HCP 7T Retinotopy Dataset by Benson, Jamison, Vu, Arcaro, et al. A PDF of the poster can be found below along with details about the broader project and links to relevant resources. Please contact Noah Benson for inquiries.
The HCP 7T retinotopy dataset is more fully documented in a paper in the Journal of Vision:
Benson NC, Jamison KW, Arcaro MJ, Vu AT, Glasser MF, Coalson TS, Van Essen DV, Yacoub E, Ugurbil K, Winawer J, Kay K (2018) The Human Connectome Project 7 Tesla retinotopy dataset: Description and population receptive field analysis. J. Vis. 18(13):23. DOI:10.1167/18.13.23
The full HCP 7T Retinotopy Dataset, including all pRF solutions for all 181 subjects, can be found at the project’s OSF site.
The Human Connectome Project, WU-Minn consortium conducted the experiments and collected the data for this project. The raw and preprocessed data can be downloaded from their database website, which requires registration but is otherwise free. The data were preprocessed using the HCP pipelines, information about which can be found here; the pipelines may be downloaded via their github page.
Tools for Interacting with HCP Retinotopy Data
The HCP retinotopy data can be accessed directly by the HCP’s workbench tool, information about which can be found here, or via neuropythy, a Python library. In particular, neuropythy can automatically download both HCP structural data and the retinotopy data and organize it into coherent Python data structures; for more information, see this page.
The retinotopic atlas (discussed on the VSS Poster) is included in the
neuropythy library, which
is publicly available on github. The raw data files that describe the
atlas (stored in FreeSurfer’s MGH format) can be found in the
neuropythy/lib/data/fsaverage/surf directory. To apply the
retinotopy atlas to a subject, we suggest one of two methods (for more
information, see this page):
- Use the neuropythy docker;
if you have Docker installed, you can simply
run the following command:
> docker run -it --rm -v "$SUBJECTS_DIR:/freesurfer_subjects" \ nben/neuropythy atlas --verbose <subject ID>
Note that you can pass the flag
--helpin place of the
--verbose <subject ID>to see further options.
- If you have Python installed, you can install the neuropythy
library using pip:
pip install neuropythy. You can then use the following command:
> python -m neuropythy atlas --verbose <subject ID>
Note that in both cases, you will need to have your
variable set to your FreeSurfer subjects directory; in the latter case you can
alternately provide an absolute or relative path the the subject’s directory in
places of the subject ID.
The VSS 2018 Poster
If your browser does not support embedded PDF content, you can download the poster here.
- The retinotopic atlas derived from the HCP 7T retinotopy dataset is an extension of previous work (Benson et al., 2014) in which a template of retinotopy was fit to group-average retinotopy data from 19 subjects; this new atlas is fit to the group-average from the 181 subjects in the HCP dataset.
- The retinotopic atlas differs from other atlases, such as the Wang et al. (2015) atlas, in that the retinotopic atlas shown here describes not only visual area boundaries of the cortical surface but also the retinotopic coordinates (polar angle, eccentricity) and the pRF size of each location in early visual cortex.
- Although the retinotopic atlas is shown with the Wang et al. (2015) atlas super-imposed for comparison, this Wang atlas was not used in the construction of the retinotopic atlas.
- The retinotopic atlas includes several regions beyond V1, V2, and V3; although these regions are shown, we have not yet systematically evaluated their accuracy and cannot recommend their use for predictive purposes. However, their inclusion in the atlas stabilizes the fitting of the V1-V3 regions, thus their inclusion improves the predictive power of these regions.
- The plot of PRF size in terms of eccentricity shows the best-fit line for each visual area as well as a shaded region that denotes the inner two quartiles of the pRF measurements. Due to the number of measurements included in these fits, the S.E.M. and 95% confidence intervals are smaller than the thickness of the lines plotted.
- Benson NC, Jamison KW, Arcaro MJ, Vu AT, Glasser MF, Coalson TS, Van Essen DV, Yacoub E, Ugurbil K, Winawer J, Kay K (2018) The HCP 7T Retinotopy Dataset. bioRxiv doi:10.1101/308247
- The WU-Minn Human Connectome Project: An overview. NeuroImage 80:62-79.
- Kay KN, Winawer J, Mezer A, Wandell BA (2013) Compressive spatial summation in human visual cortex. J Neurophysiol 110:481-94.
- Wang L, Mruczek RE, Arcaro MJ, Kastner S (2015) Probabilistic Maps of Visual Topography in Human Cortex. Cereb Cortex 5:3911-31.
- Benson NC, Butt OH, Brainard DH, Aguirre GK (2014) Correction of Distortion in Flattened Representations of the Cortical Surface Allows Prediction of V1-V3 Functional Organization from Anatomy. PLOS Comput Biol 10(3):e1003538