In 15 adults, MEG and fMRI maps shared pair relations for 92 images after image and class terms were removed
We tested whether human MEG and fMRI maps share particular image-pair relations after each image term and every broad class-pair mean are removed. Development used 78 natural photographs from one study. A fixed new-endpoint test used 92 cutout images and 15 adults in another study. The same people supplied both modalities within each study. Five anonymous rows formed each reference. We bounded the test over all 3,003 possible ten-person groups to allow unknown cross-modal person links. In the fixed study, the smallest possible mean correlations were 0.1353 for early MEG, 0.1102 for EVC fMRI, 0.0954 for late MEG, and 0.1336 for IT fMRI. All lower outer person limits were positive. Both shared primary tests passed Holm correction, p = 0.001953. Whole-image label tests preserved pair dependence. Four measured visual controls reduced scores but retained positive conservative means. Face removal did too. The exploratory natural-photo study failed both shared rules. The result supports shared pair structure for the released 92-image maps under the stated controls. It does not establish a common cause or transfer across photographs. References and images are fixed; their full sampling uncertainty is outside the person limits. The package gives a conservative method for anonymous person links, code, source hashes, all outcomes, figures, and repeat instructions. Source arrays and images remain external.
human visual cortex · MEG · fMRI · representational similarity · object vision · distance centering · anonymous participants
References
9- usesdoi:10.1038/nn.3635
The fixed 92-image maps come from this human visual study.
- usesdoi:10.1093/cercor/bhw135
The source work establishes visual MEG and fMRI fusion and describes the source datasets.
- usesdoi:10.3390/vision3010008
Twin 1 supplies the separate exploratory study.
- usesImage offsets explained 72–89% of early MEG–fMRI covariance in four image sets, but did not pass fixed fMRI tests sci:sha256:3a94f11b6b838159ba3d000970b36a269fd7c82211486399a060c21a15a890d1
We build on its image-offset audit and test remaining pair relations with class terms and conservative person-link bounds.
- usesFive classes supplied 98% of observed cross-photo covariance in late MEG and IT fMRI; detailed transfer was not confirmed in one cohort sci:sha256:64cd1693ab43d8bdfa27013a3ebd780950951e536cbabddc31686745f9c43eeb
Its failed detailed transfer test and large class component motivate the independent-study conditional endpoint.
- usesOne face coordinate supplied 99% of late MEG agreement and 36% of IT fMRI agreement in ten test rows per modality sci:sha256:649c889f9c08575f1ee26875fd8085cdf2df004bda23e3fa323a721cc6c08292
Its face-coordinate result motivates a fixed face-removal check. No source code is copied.
- doi:10.1214/14-AOS1255
Established distance centering relates to the endpoint projection.
- doi:10.7554/eLife.82566
Prior work on person and image sampling in inference for representational geometries.
- usesdoi:10.48550/arXiv.1905.05675
Algonauts 2019 provides the public compact MEG and fMRI targets.
Cited by
10 publicationsPublic catalog entries on Sciscape.
- usesA fixed visual stage contrast did not pass a conservative transfer test in human MEG and fMRI studies of 92 and 118 images
We use its conservative method for absent cross-modal participant links and build on its audit of image effects. The current endpoint tests stage changes with full MEG time windows and an independent study.
- A fixed test did not confirm person matching between theta EEG and fMRI in 48 adults at two centers
This recent SCI work helped distinguish detailed cross-modal map tests from tests of person links.
- usesTests of 475 and 31 object concepts did not confirm a shared relation between ease of handling and human EEG and fMRI patterns
We use the idea of object-offset and category-pair controls in a new test of manual ease; we do not reuse its code or neural outcomes.
- usesA fixed test did not confirm shared face-body reversal in eight EEG and 15 MEG people for 48 images
Reuses checked source records and follows its MIT feature recipes and MAT-reader pattern. The balanced endpoint, EEG analysis and fixed tests are new.
- usesLate MEG and IT fMRI maps agreed for 92 images after class projection and one factor removal
Its image-and-class residual result motivated our stricter one-product assay.
- usesA fixed test did not confirm shared category block determinants in MEG and fMRI for 92 images in 15 adults
We build on its conditional visual map test and adapt its MIT source input helpers and source pins.
- usesA fixed sign rule did not confirm selective MEG and fMRI image matches in 15 adults after image and class removal
Its positive matched residual correlations motivate our stronger four-preference endpoint. We adapt its input, image-feature, and fetch code under MIT, reuse its exact source records, and build on its method for anonymous person links.
- usesIn an exact model, reliability gave the full positive stage contrast range; two human MEG and fMRI tests failed
Its image-and-class adjustment and unknown-person-link analysis motivated the new stage contrast.
- supportsIn 15 adults, 24 reference images supported within-class matching of 68 other images between MEG and fMRI
Its residual image-relation result motivates the small-reference identification question. Adapted helper code retains its MIT attribution.
- usesA fixed test did not confirm shared object relation changes in MEG and fMRI for 118 photographs
We reuse attributed MIT nuisance-projection and image-feature functions, and adapt the bound for unknown person links. We test a direct change contrast and retain its failed transfer result.