Overview -------- This project repository contains all the analysis code and datasets associated with the following manuscript (currently under review): **Vinke, Bloem, & Ling (***under review***). Saturating nonlinearities of contrast response in human visual cortex.** Questions/Comments: vinke [at] bu [dot] edu **Abstract:** Response nonlinearities are ubiquitous throughout the brain, especially within sensory cortices where changes in stimulus intensity typically produce compressed responses. Although this relationship is well-established in electrophysiological measurements, it remains controversial whether the same nonlinearities hold for population-based measurements obtained with human fMRI. We propose that these purported disparities are not contingent upon measurement type, and are instead largely dependent upon the visual system state at the time of interrogation. We show that deploying a contrast adaptation paradigm permits reliable measurements of saturating sigmoidal contrast response functions (10 participants, 7 female). When not controlling the adaptation state, our results coincide with previous fMRI studies, yielding non-saturating, largely linear contrast responses. These findings highlight the important role of adaptation in manifesting measurable nonlinear responses within human visual cortex, reconciling discrepancies reported in vision neuroscience, re-establishing the qualitative relationship between stimulus intensity and response across different neural measures, and the concerted study of cortical gain control. ---------- Experiment 1 Dataset -------- *Exp1_Data.zip* - **S001 ... S008** [directory] - **ROImasks** [directory] - v[*#* ]-PRFmask.nii.gz Volumetric ROI masks for areas V1, V2, and V3. - **PRFvols** [directory] - [*?* ]h.pRF_[*ang ecc R2 rfsize* ].VOL.nii.gz Volumetric population receptive field results for V1/V2/V3, separated by cortical hemisphere: ang = Polar angle (radial degrees) ecc = Eccentricity (degrees of visual angle) R2 = PRF model goodness-of-fit (R-squared) rfsize = Receptive field size - **FIR_analysis** [directory] - c[*1..9* ].ces.nii.gz Volumetric deconvolution results for each stimulus contrast condition (1 = 2.7%, 2 = 4%, 3 = 5.3%, 4 = 8%, 5 = 16%, 6 = 32%, 7 = 48%, 8 = 64%, 9 = 96% Michelson Contrast). - fnsr.nii.gz Voxel-wise functional SNR calculation. - mask.nii.gz Simple binary volumetric mask. - **template.nii.gz** Template volume used for saving out new volumes created during analyses. - **voxCR_fitFIRmax_Exp1.v[#].nii.gz** Output from *voxelSelection_CRestimates.m* containing voxel-wise Gaussian goodness-of-fit values. - **voxSelect_fitFIRmax_Exp1.v[#].nii.gz** Output from *voxelSelection_CRestimates.m* containing voxel-wise contrast response estimates. - **allFIR_fits_wAdapt_GROUP_CRF_09-Oct-2021.mat** Experiment 1 (adaptation) output from *FIRanalysis_VoxelWise.m* containing ROI and voxel-wise model fitting results. ---------- Experiment 2 Dataset -------- *Exp2_Data.zip* - **S002 ... S010** [directory] - **ROImasks** [directory] - v[*#* ]-PRFmask.nii.gz Volumetric ROI masks for areas V1, V2, and V3. - **PRFvols** [directory] - [*?* ]h.pRF_[*ang ecc R2 rfsize* ].VOL.nii.gz Volumetric population receptive field results for V1/V2/V3, separated by cortical hemisphere: ang = Polar angle (radial degrees) ecc = Eccentricity (degrees of visual angle) R2 = PRF model goodness-of-fit (R-squared) rfsize = Receptive field size - **FIR_analysis** [directory] - c[*1..9* ].ces.nii.gz Volumetric deconvolution results for each stimulus contrast condition (1 = 2.7%, 2 = 4%, 3 = 5.3%, 4 = 8%, 5 = 16%, 6 = 32%, 7 = 48%, 8 = 64%, 9 = 96% Michelson Contrast). - fnsr.nii.gz Voxel-wise functional SNR calculation. - mask.nii.gz Simple binary volumetric mask. - **template.nii.gz** Template volume used for saving out new volumes created during analyses. - **voxCR_fitFIRmax_Exp2.v[#].nii.gz** Output from *voxelSelection_CRestimates.m* containing voxel-wise Gaussian goodness-of-fit values. - **voxSelect_fitFIRmax_Exp2.v[#].nii.gz** Output from *voxelSelection_CRestimates.m* containing voxel-wise contrast response estimates. - **allFIR_fits_NoAdapt_GROUP_CRF_09-Oct-2021.mat** Experiment 2 (no adaptation) output from *FIRanalysis_VoxelWise.m* containing ROI and voxel-wise model fitting results. ---------- Data Analysis scripts -------- *Analysis_Scripts.zip* Zip file containing custom MATLAB scripts (plus sub-functions) to run voxel selection, contrast estimation, and model fitting analyses. - **voxelSelection_CRestimates.m** MATLAB script for fitting a Gaussian function to the ROI-average and voxel-wise 96% contrast deconvolution results for all ROIs (V1 - V3). The voxel-wise Gaussian goodness-of-fit values are saved out for each subject and ROI (per experiment) as: - voxSelect_fitFIRmax_Exp[*#* ].v[*#* ].nii.gz This script also computes the voxel-wise contrast response estimate for all contrast levels (2.7% - 96%) by averaging the deconvolution beta weights within a fixed post-stimulus time window. The voxel-wise contrast response estimates are saved out for each subject and ROI (per experiment) as: - voxCR_fitFIRmax_Exp[*#* ].v[*#* ].nii.gz - **FIRanalysis_VoxelWise.m** MATLAB script to perform Linear and Naka-Rushton model fitting (ROI and voxel-wise) after further constraining voxel selection based on population receptive field results. All results are saved out as a .mat file. - allFIR_fits_[*??* ]Adapt_GROUP_CRF_[*DATE* ].mat ***FIR(subj,roi).avgFIR*** = ROI-average contrast response estimates ***FIR(subj,roi).std_FIR*** = StDev of ROI-average contrast response estimates ***FIR(subj,roi).allCRF*** = Voxel-wise contrast response estimates ***FIR(subj,roi).avgCRF*** = NR function fit (ROI-level) ***FIR(subj,roi).errDiff_NR*** = NR model fit residuals (ROI-level) ***FIR(subj,roi).errDiff_Lin*** = Linear model fit residuals (ROI-level) ***FIR(subj,roi).R2_allVoxels*** = NR model goodness-of-fit (voxel-wise) ***FIR(subj,roi).R2_avgFIR*** = NR model goodness-of-fit (ROI-level) ***FIR(subj,roi).rsq_voxelSelection*** = Final voxel selection mask ***FIR(subj,roi).linearCoordinates*** = Volume coordinates for final voxel selection ***FIR(subj,roi).sse_allVoxels*** = NR model fit residuals (voxel-wise) ***FIR(subj,roi).sse_lin*** = Linear model fit residuals (voxel-wise) ***FIR(subj,roi).R2_lin*** = Linear model goodness-of-fit (voxel-wise) ***FIR(subj,roi).AIC_lin*** = Linear model Akaike’s information criterion (voxel-wise) ***FIR(subj,roi).AIC_CRF*** = NR model Akaike’s information criterion (voxel-wise) ***FIR(subj,roi).AIC_linROI*** = Linear model Akaike’s information criterion (ROI-level) ***FIR(subj,roi).AIC_CRF_ROI*** = NR model Akaike’s information criterion (ROI-level) ***FIR(subj,roi).R2_wholeROI*** = Linear model goodness-of-fit (ROI-level) ***ALL_estParams(subj, roi).wholeROI_params*** = NR model parameter estimates (ROI-level) ***ALL_estParams(subj, roi).wholeROI_betas*** = Linear model parameter estimates (ROI-level) ***ALL_estParams(subj, roi).est_params_allVoxels*** = NR model parameter estimates (voxel-wise) ***ALL_estParams(subj, roi).beta_allVoxels*** = Linear model parameter estimates (voxel-wise) - **NRFunc.m** MATLAB sub-function defining Naka-Rushton function used for model fitting and plotting. - **sseval.m** MATLAB sub-function for evaluating SSE used during fitting procedures. ---------- Stimulus Presentation Code ---------------- *Stimulus_Code.zip* Zip file containing MATLAB/PsychToolbox script required to present stimuli used in this study. Some small changes may be necessary depending on your local setup and equipment. This code is setup to present stimuli with (Experiment 1) or without (Experiment 2) an initial contrast adaptation period. - **CRF_wContrastPetals_cortMag.m** Main script to run the experiment. - **Paradigm Files** These files were initially generated using the [optseq2][1] routine, and then edited slightly to incorporate the initial adaptation period. ***p_files_wAdapt/block#.par*** Sub-directory containing paradigm files used to determine stimulus presentation timing for each fMRI acquisition run in Experiment 1. ***p_files_noAdapt/block#.par*** Sub-directory containing paradigm files used to determine stimulus presentation timing for each fMRI acquisition run in Experiment 2. [1]: https://surfer.nmr.mgh.harvard.edu/fswiki/optseq2