RetinotopicMapping
Retinotopic mapping of human visual cortex
Contents |
This page provides a detailed description (incl. stimulus files and analysis scripts) of retinotopic mapping using two kinds of stimuli: 1) a 24-region multifocal stimulus (multifocal mapping) and 2) a blocked presentation of object stimuli at nine different visual field positions (object mapping). We use these stimuli routinely for retinotopic mapping of the human visual cortex, and the methods are intended to be easy to use. The multifocal mapping is optimal for visual areas V1, V2, and V3, but typically performs okay also for visual areas hV4 and V3AB. The object mapping is well-suited for the retinotopic mapping of visual areas V1-3, hV4, V3AB, IPS0, LO1, LO2, VO1, and depending on the quality of the data, also for example of VO2 and TO maps. Both mapping methods can also be used as functional localizers of multiple retinotopic regions-of-interest.
Software requirements for retinotopic mapping:
- Presentation (for stimulus presentation)
- Matlab with SPM8 toolbox (for data analysis)
- Freesurfer (for visualization of the retinotopic maps)
If you use our methods in your research, please cite the article:
Henriksson L, Karvonen J, Salminen-Vaparanta N, Railo H, and Vanni S. Retinotopic Maps, Spatial Tuning, and Locations of Human Visual Areas in Surface Coordinates Characterized with Multifocal and Blocked fMRI Designs. PLoS ONE, 7(5): e36859, 2012. doi:10.1371/journal.pone.0036859.
If you have questions or comments, please contact: Linda Henriksson
See also: A surface-based probabilistic atlas of the retinotopic visual areas
MULTIFOCAL MAPPING
Stimuli
Download the stimulus images, stimulus setup files and Presentation scenarios: MF24.zip
Our data are typically acquired with echo planar imaging using single-shot gradient-echo sequence with the following imaging parameters
| parameter | TR | no. of time points | no. of slices | slice thick. | FOV | matrix | TE | flip angle |
|---|---|---|---|---|---|---|---|---|
| value | 1.8 s | 132 | 29 | 3.0 mm | 20 cm | 64 x 64 | 30 ms | 60 deg |
If you use different TR or number of slices, remember to modify the stimulus scenario files!
Preprocessing & regressor estimation of the data with SPM8
For help with the use of SPM8, see SPM8 Manual
Preprocessing of the multifocal data with SPM8 using a batch script
- Batch script for preprocessing of the multifocal data with SPM8: conv_sli_rea_mfdata_example.m, includes:
- DICOM to NIFTI format conversion for functional and anatomical data
- slice timing correction
- movement correction (realign and reslice)
- option for spatial smoothing
Statistical analysis of the multifocal data with SPM8 using a batch script
- Batch script for statistical analysis of the multifocal data with SPM8: estimate_ANY_MF24_ANY_rp.m, includes:
- desing matrix setup
- estimation of the parameters
- contrast vector setup
Visualizing the multifocal results with SPM8
Data from SPM8 to Freesurfer
For help with the use of Freesurfer, see: FreeSurferWiki
1. Convert the high resolution anatomical data (the anatomical data from which you have reconstructed the Freesurfer surfaces) to NIfTI format with Freesurfer* **:
- mri_convert orig.mgz anat.nii
(* anatomical data in path $SUBJECTS_DIR/Subject/mri)
(** needs to be done only once for each subject, this is the reference volume for coregistration)
2. Copy the new data (eg. *layer.img-files, spmT-maps, mean functional image, low resolution anatomical image) to Freesurfer subject directory, something like $SUBJECTS_DIR/Subject/functional/examnumber)
3. Coregister with SPM8:
- Coregister (Estimate)
- Reference image: anat.nii
- Source image: low resolution anatomical (measured with functional data)
- Other images: everything else you copied to $SUBJECTS_DIR/Subject/functional/examnumber
4. Create register.dat file in Freesurfer: tkregister2 --mov mean*.img --s Subject --regheader --noedit --reg register.dat
5. Check Coregistration in Freesurfer: tkregister2 --mov mean*.img --s Subject --reg register.dat
Constructing the retinotopic maps with matlab
batch_spm8_pol_sc_mf_all_basic_rp.m
spm8_polar_sc_analysis_basic.m
Visualizing the retinotopic maps with Freesurfer
For help with the use of Freesurfer, see: FreeSurferWiki
Visualize the eccentrity and polar angle maps with Freesurfer
1. tksurfer Subject rh inflated
2. File -> Load Overlay
- Load Overlay: *layer1.img, Specify registration file: register.dat, Into Field: Overlay Layer 1 (empty), [OK]
3. File -> Load Overlay
- Load Overlay: *layer2.img, Specify registration file: register.dat, Into Field: Overlay Layer 2 (empty), [OK]
4. View -> Configure -> Overlay
- Set the values as shown in the figures below
OR use a tcl-file to load the data and configure the overlay
1. tksurfer Subject rh inflated -tcl load_mfdata.tcl
OBJECT MAPPING
Stimuli
Download the stimulus images, stimulus setup files and Presentation scenarios: OBJ9.zip
The objects in the stimuli were extracted from photographs obtained from free online photograph libraries (www.freeimages.co.uk and www.morguefile.com).
Our data are typically acquired with echo planar imaging using single-shot gradient-echo sequence with the following imaging parameters
| parameter | TR | no. of time points | no. of slices | slice thick. | FOV | matrix | TE | flip angle |
|---|---|---|---|---|---|---|---|---|
| value | 1.8 s | 132 | 29 | 3.0 mm | 20 cm | 64 x 64 | 30 ms | 60 deg |
If you use different TR or number of slices, remember to modify the stimulus scenario files!
Preprocessing & regressor estimation of the data with SPM8
For help with the use of SPM8, see SPM8 Manual
Preprocessing of the data with SPM8 using a batch script
- Batch script for preprocessing of the object data with SPM8: conv_sli_rea_lodata_example.m, includes:
- DICOM to NIFTI format conversion for functional and anatomical data
- slice timing correction
- movement correction (realign and reslice)
- option for spatial smoothing
Statistical analysis of the data with SPM8 using a batch script
- Batch script for statistical analysis of the object data with SPM8: estimate_ANY_LO_ANY_rp.m, includes:
- desing matrix setup
- estimation of the parameters
- contrast vector setup
Visualizing the multifocal results with SPM8
Data from SPM8 to Freesurfer
1. Convert the high resolution anatomical data (the anatomical data from which you have reconstructed the Freesurfer surfaces) to NIfTI format with Freesurfer* **:
- mri_convert orig.mgz anat.nii
(* anatomical data in path $SUBJECTS_DIR/Subject/mri)
(** needs to be done only once for each subject, this is the reference volume for coregistration)
2. Copy the new data (eg. *layer.img-files, spmT-maps, mean functional image, low resolution anatomical image) to Freesurfer subject directory, something like $SUBJECTS_DIR/Subject/functional/examnumber)
3. Coregister with SPM8:
- Coregister (Estimate)
- Reference image: anat.nii
- Source image: low resolution anatomical (measured with functional data)
- Other images: everything else you copied to $SUBJECTS_DIR/Subject/functional/examnumber
4. Create register.dat file in Freesurfer: tkregister2 --mov mean*.img --s Subject --regheader --noedit --reg register.dat
5. Check Coregistration in Freesurfer: tkregister2 --mov mean*.img --s Subject --reg register.dat
Constructing the retinotopic maps with matlab
batch_spm8_pol_sc_lo_all_basic_rp.m
spm8_polar_sc_analysis_basic.m
Visualizing the retinotopic maps with Freesurfer
Visualize the eccentrity and polar angle maps with Freesurfer
1. tksurfer Subject rh inflated
2. File -> Load Overlay
- Load Overlay: *layer1.img, Specify registration file: register.dat, Into Field: Overlay Layer 1 (empty), [OK]
3. File -> Load Overlay
- Load Overlay: *layer2.img, Specify registration file: register.dat, Into Field: Overlay Layer 2 (empty), [OK]
4. View -> Configure -> Overlay
- Set the values as shown in the figures below
OR use a tcl-file to load the data and configure the overlay
1. tksurfer Subject rh inflated -tcl load_lodata.tcl
Delineation of the Retinotopic Visual Areas
Here is an example of the delination of the retinotopic visual areas based on the object mapping data. First we examine the retinotopic maps in the low-level visual areas and then we move to higher-level visual areas.
V1, V2, V3
V1 is located within the calcarine sulcus and represents the full contralateral (opposite) visual hemifield. The upper visual field is mapped approximately onto the lower lip of the calcarine sulcus and the lower visual field onto the upper lip of the calcarine sulcus. The fovea (central retina) is represented on the occipital pole, and the eccentricity map runs in posterior-anterior direction (fovea-peripheral visual field).
V1, dorsal division of V2 (V2d), dorsal division of V3 (V3d)
The representation of the lower vertical meridian defines the border between V1 and dorsal division of area V2 (V2d). V2d represents the lower quadrant of the contralateral hemifield. The representation of the horizontal meridian defines the border between V2d and dorsal division of area V3 (V3d). V3d also represents the lower quadrant of the contralateral hemifield.
V1, ventral division of V2 (V2v), ventral division of V3 (V3v)
The representation of the upper vertical meridian defines the border between V1 and ventral division of area V2 (V2v). V2v represents the upper quadrant of the contralateral hemifield. The representation of the horizontal meridian defines the border between V2v and ventral division of area V3 (V3v). V3v also represents the upper quadrant of the contralateral hemifield.
hV4, VO1 VO2
The representation of the upper vertical meridian define the border between V3v and retinotopic maps in the ventral occipital cortex. According to the current view, area hV4 represents the full contralateral hemifield. In hV4, the eccentricity map runs parallel to the maps in V1-V3 (confluent representation of the fovea), but is shorter than maps in V1-V3 (outer boundary defined by the representation of the peripheral visual field).
The eccentricity maps in ventral occipital (VO) areas run lateral-medial direction (see the distinct representation of the fovea!). The representation of the upper vertical meridian defines the border between VO1 and VO2. The representation of the lower vertical meidian defines the outer boundary of VO2. The tuning strength map can assist the identification of the ventral occipital areas from the loosely defined pFUS region, which does not show a consistent retinotopic organization across subjects.
V3AB, IPS0, LO1, LO2
The representation of the lower vertical meridian defines the border between V3d and V3AB. Area V3AB represents the full contralateral hemifield, and the representation of the upper vertical meridian defines the border between V3AB and IPS0. Area IPS0 also represents the full contralateral hemifield.
The representation of the lower vertical meridian defines the border between V3AB and LO1. Area LO1 typically shares its posterior border also with area V3d. LO1 and LO2 both represent the contralateral hemifield and the representation of the upper vertical meridian defines the border between LO1 and LO2.