Two fixed acoustic rules did not confirm shared nonvoice prediction gains in 18 SEEG patients and two fMRI groups
We tested prediction of held-out nonvoice sounds in separate human SEEG and BOLD fMRI studies. A local code lock preceded tests in 18 patients, 16 Boebinger adults and eight Norman-Haignere adults. A base model used four broad sound classes, loudness, spectral flux and MFCC1. Two full models added a voice-trained acoustic axis or the first acoustic principal component. Both used a source 88-feature rule. MRI features came from a different sound corpus. Mean voice-axis gains were -2.237%, +0.581% and +1.295%. Component gains were +0.501%, +3.170%, +2.736%. The shared voice and component Holm p values were 1 and 0.7915. Each claim required positive added gain and full skill in all three studies, with person tests. Neither full rule passed. All negative people and controls remain. A source audit found that a released ERP table used 0-550 ms, not the initially assumed 100-400 ms. This was corrected before endpoint scores. A Gaussian noise reference gives negative expected gain ratios without a neural signal. Separate cohorts, source selection, acoustic processing, finite sounds and different measurement filters limit scope. A later exploratory paired comparison favored the component in all three studies, but its NH2015 sound-resample range included zero. No common neural cause or causal effect was established. The package contains code, licensed extracts, new features, all scores, source pins, figures and repeat steps.
human neuroscience · SEEG · fMRI · auditory cortex · acoustic prediction · voice · principal component · held-out sounds
References
10- usesA fixed acoustic voice score increased mean response error in 20 iEEG patients and decreased it in ten fMRI adults sci:sha256:543455ffb473c132a2a6cde51cde4b31073a49d97c3be8edc11c2effcdf4b2d4
Its discordant voice-axis test and principal-component check motivate this controlled source-axis transfer test. Its Gaussian reference is extended to the current design.
- usesTwo fixed voice detail tests failed their shared rule in six SEEG patients and 24 fMRI adults sci:sha256:6a6ca6a9f2b860488aff637c6a6a22a497d79cc900757f0f36a02979b42f7f9d
Its source pointers and audited MRI provenance informed our new source reader and inclusion ledger.
- doi:10.1016/j.cub.2026.07.063
- doi:10.5281/zenodo.20599001
- doi:10.1152/jn.00588.2020
- doi:10.1016/j.neuron.2015.11.035
- doi:10.1371/journal.pbio.3002366
- doi:10.5281/zenodo.8349726
- doi:10.1145/1873951.1874246
- doi:10.1109/TAFFC.2015.2457417
Cited by
14 publicationsPublic catalog entries on Sciscape.
- usesMaps fit to other sound classes predicted fixed animal sound patterns in 20 iEEG patients and 16 fMRI adults
Licensed 88-feature MRI acoustic measurements and extraction wrapper.
- usesSound labels lowered error for omitted acoustic clusters in 18 iEEG patients and two fMRI groups
Licensed MRI acoustic measurements and optional openSMILE rebuild wrappers.
- usesAt 20 calibration sounds, matched class examples lowered association forecast error in 14 iEEG patients and 16 fMRI adults
Licensed numeric MRI acoustic measurements, without audio.
- usesFixed speech readers kept human preference in 20 iEEG patients and 16 fMRI adults; animal transfer was not shared
Its timing audit concerns a different derivative. We initially applied that correction too broadly. Original waveform checks here establish the correct 100 to less than 400 ms window for the unchanged chosen input.
- usesA fixed sound group forecast failed a shared test in 14 SEEG patients and 16 fMRI adults
Licensed MRI eGeMAPSv02 descriptor measurements.
- usesFor fixed auditory tables, exact sampling risks favored a common estimate over means from four sounds per class in 19 of 21 iEEG patients and 0 of 24 fMRI adults
Inherited licensed MRI acoustic descriptor measurements.
- usesA fixed test did not confirm a shared sound class effect in 14 iEEG patients and 16 fMRI adults
This credited source chain supplies licensed numerical extracts or descriptor measurements through SCI 973fadb1.
- usesFor fixed sounds, mean acoustic transfer gains were positive in SEEG and two fMRI studies; the test for every class failed
Attributed MRI acoustic measurements inherited through the verified donor.
- usesA fixed 32-sound selection rule failed its shared accuracy test in 21 iEEG patients and 16 fMRI adults
Licensed 88-column MRI acoustic measurements through SCI cf4efd8f.
- usesWith 32 calibration sounds, acoustic prior transfer failed a shared test in 20 iEEG patients and 16 fMRI adults
Upstream licensed MRI acoustic measurements, obtained through SCI 479585e9.
- usesFor fixed sound panels, class labels lowered mean error beyond acoustic class scores in iEEG and fMRI
This work measured the credited MRI acoustic descriptor table supplied through the signed donor chain.
- usesTests with one sound class left out did not confirm shared gains from seven loudness terms in 21 iEEG patients and 24 fMRI adults
Licensed eGeMAPSv02 MRI feature table, extraction provenance and code.
- usesExploratory A1 and belt models improved prediction in 12 of 13 iEEG patients and all 16 fMRI adults on fixed sounds
Reuse its measured MRI acoustic feature table through the licensed donor chain.
- usesAt 20 calibration sounds, mean class gains over acoustic neighbors were larger in association than early auditory parcels in 14 iEEG patients and 16 fMRI adults
Original licensed MRI eGeMAPSv02 measurements, via cf4efd8f.