A fixed test did not confirm shared finger pair structure in five ECoG patients and seven fMRI adults
We tested whether a finger distance template transfers between human electrical and fMRI data. Three ECoG patients set the template. Five other patients and seven fMRI adults supplied the fixed test. The cohorts came from separate public studies. One source record failed the cue rule. The plan fixed raw agreement and agreement after single-finger and finger-order terms were removed. Raw correlations were positive in all test people. Mean raw correlations were 0.544 in ECoG and 0.776 in MRI. Their fixed-template 95% person intervals were 0.371 to 0.692 and 0.723 to 0.829. The raw shared Holm p was 0.0625. After the source exclusion, this two-claim sign rule could not reach 0.05. Adjusted means were -0.258 in ECoG and 0.772 in MRI. Their intervals were -0.698 to 0.182 and 0.676 to 0.852. The adjusted shared p was 0.8125. Neither fixed shared claim passed. Template resampling also made raw ECoG agreement uncertain. A constructed example had raw correlation 0.983 and exactly opposite adjusted pair relations. The contribution is a controlled benchmark, an algebraic check, and full numerical results. Strong MRI agreement did not establish electrical repetition or cross-modal generalization. Small samples, clinical coverage, task differences, source processing and five fixed fingers limit scope. The package preserves code, source hashes and terms, plans, all scores, uncertainty draws, figures and repeat instructions. Neural inputs stay external.
finger representation · ECoG · fMRI · representational geometry · cross-modal validation · human motor cortex · negative result
References
6- usesdoi:10.1371/journal.pcbi.1002655
This task study is the source of the human finger-flexion ECoG recordings.
- usesdoi:10.1038/s41562-019-0678-3
The public library supplies the original archive, task notes and source terms.
- usesdoi:10.1038/nn.4038
The seven-person 7 T single-finger study supplies the processed fMRI test cohort. Its known finger geometry informed the question.
- usesdoi:10.1016/j.neuroimage.2017.08.051
The PCM tutorial supplies processed M1 matrices with finger and run labels.
- usesTests in seven ECoG patients and eight fMRI adults did not establish shared coding of individual stimuli sci:sha256:e4b2933c48c9d2368446d499ec0883dbc45cfa770a5c419eb21a7e69dafbaab2
Its sequence-confound audit informed our cue-position audit and limits on detailed response claims.
- usesdoi:10.1371/journal.pcbi.1005508
Representational model methods support comparison of condition-distance structures across distinct feature spaces.
Cited by
5 publicationsPublic catalog entries on Sciscape.
- usesA shared finger profile test failed in six ECoG patients and six fMRI adults; unequal errors gave 50.5% mean matching in a model
We build on its source manifest, data access code and trial boundary logic. Its failed shared distance-template test motivates a different record matching endpoint.
- usesA fixed test did not confirm greater thumb contrast energy in two ECoG patients and six fMRI adults
It supplied upstream access code and a related failed adjusted geometry-transfer benchmark.
- usesWith stable means and independent repeat errors, mean finger response area was positive in five ECoG patients and six fMRI adults after feature mean removal
It supplied upstream source access and trial-boundary code and a prior finger-distance benchmark.
- usesFixed tests did not confirm four-axis finger gains in four ECoG patients and six fMRI adults
We use its source manifest and access code. Its failed detailed finger-map test informs this different endpoint.
- A five-finger order test with single-finger terms removed did not confirm a shared pattern in six ECoG patients and six fMRI adults
Its failed detailed finger-template test motivated a fixed test of the order term itself. We preserve its inherited code credit.