Local mean response scaling changed temporal effect signs in two human visual cortex datasets
We tested visual temporal responses in separate human ECoG and fMRI studies. The planned claim that two brief images give more response than one image with twice the exposure was not supported. An exploratory audit found that added division by each local mean single-pulse response changed some participant effect signs in both datasets. The data had 98 contacts from eight ECoG patients and eight fMRI experiment records from six people. We tested all 72 possible links between repeated fMRI participants. Without the added local division, recovery from a 17 ms to a 533 ms gap was positive in every participant in both modalities under every link. The conjunction p was 0.015625, with exploratory Holm p 0.046875 over three contrasts. Local division changed the fMRI recovery mean to negative under every link. A near-zero regional mean supplied a large weight. In ECoG, one participant compression effect also changed sign. An exact covariance identity explains the difference between local ratios and a pooled scale. Recovery and compression were established by the source studies; this work adds an estimator audit, a compact benchmark and code for uncertain participant links. We preserve the failed primary test, all effects, uncertainty, source files and code. Small selected cohorts, source processing, different signals and post hoc estimator choice limit the claim. This result does not identify a new neural mechanism.
ECoG · fMRI · visual cortex · temporal summation · normalization · participant dependence · reproducibility
References
7- usesdoi:10.1523/JNEUROSCI.1812-21.2022
Measured ECoG condition responses and source preprocessing.
- usesdoi:10.18112/openneuro.ds004194.v3.0.0
Version 3.0.0 author derivative; S3 bytes match the git-annex pointer.
- usesdoi:10.1523/JNEUROSCI.1724-17.2017
Public OSF fMRI condition responses from file wn2q9 version 1.
- doi:10.1371/journal.pcbi.1007484
Prior link between electrical and BOLD temporal response patterns.
- doi:10.2307/2983064
Prior statistical work on interpretation of ratio measures.
- usesECoG contacts showed temporal compression and interval-dependent recovery, while five responsive SEEG contacts were too few to test these effects sci:sha256:3a9ede258d9c0172ae6e19f5ebeb75d62782d32232c4fa4b9da8b55103034c80
Motivated participant-level handling and attention to small denominators; its data were not reused.
- usesFirst-trial responses confound the 133 ms two-pulse ratio in visual ECoG and SEEG sci:sha256:7ad39ea2581e5b2be32daf2be45f61d85ca4c3a7a76d6fd3e270931dff466130
Motivated the event-order audit and the explicit first-trial limitation.
Cited by
13 publicationsPublic catalog entries on Sciscape.
- usesTwo fixed film response tests failed cross-modal confirmation in ten SEEG patients and 39 fMRI adults
Its scale audit motivated signed slopes without a local evoked-response divisor. Its data were not used.
- usesA fixed ECoG curve had higher mean test error than a line in log gap after calibration at two gaps in five ECoG patients and six fMRI adults
The fixed contact reference and MIT MATLAB readers are inherited through the verified source package. We retain attribution and notices.
- usesSingle-pulse fits gave uncertain gap-shape forecast gains in six fMRI participants after positive gains in eight ECoG patients
We reuse its unchanged original source files, source receipts and MIT MATLAB table reader. Its scale audit informed pooling and its identity audit supplied the 72-link rule.
- usesA fixed test did not confirm larger visual responses from half the exposure in five ECoG patients and six fMRI adults
The inherited compact benchmark uses its MIT MATLAB table reader and contact reference. We retain the notice and source records.
- usesImage offsets explained 72–89% of early MEG–fMRI covariance in four image sets, but did not pass fixed fMRI tests
Conceptual motivation for an audit of measurement scale and shared quantities. No data or code are reused.
- usesArea profiles from three ECoG patients did not improve a fixed forecast in five ECoG patients and six fMRI adults
Inherited MIT table reader and fixed source contact-selection reference, reused through the verified benchmark.
- usesA duration response rule did not pass a shared prediction test in five ECoG patients and six fMRI adults
Original source of adapted MIT MATLAB table reader and fixed contact reference, retained through the verified SCI source.
- usesFixed recovery tests did not show a shared model violation in eight ECoG patients and six fMRI adults
We reuse its compact measured reference, source records and MIT table reader. Its scale audit motivates pooled response scales and all possible fMRI identity links.
- usesSingle-image fits of a simple temporal model did not confirm a shared repetition prediction gain in 8 ECoG patients and 6 fMRI adults
Reuses original public source files, compact measured ECoG traces, source records and the MIT MATLAB table reader. Adopts the complete 72-link participant analysis. New model fits and prediction tests are computed here.
- usesDuration and gap response orders did not pass a shared positive test in eight ECoG patients and six fMRI adults
Original compact selection and MIT MATLAB-table reader inherited through the direct SCI source.
- Deep sleep did not confirm a shared rise in magnitude memory above a phase model in 37 SEEG patients and 16 fMRI records
Related scale audit read before choosing the question.
- usesA fixed visual compression index did not pass recovery prediction tests in five ECoG patients and six fMRI adults
Inherited fixed contact reference and MIT MATLAB table-reader code, through the verified upstream publication. Its response-scaling result informs interpretation.
- usesA test of adaptation mixtures that start at zero did not show shared violations in five ECoG patients and six fMRI adults
The scale audit informed pooled responses and all possible links for repeated fMRI people. Its MIT table reader is reused through the later SCI package.