At 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
We tested whether class examples improve sound response forecasts more in association than early auditory parcels. Separate public studies supplied signed iEEG field means and BOLD fMRI. Four people per study supplied development. A local code lock preceded new person scores in 14 other patients and 16 other adults. Each model used 20 calibration sounds, four per class. One used same-class examples. The other used four acoustic neighbors. Both used fixed shrinkage. Mean regional gain differences were 1.966 and 6.927 reference-error points. Their 95% person Student limits were [0.673,3.260] and [6.162,7.692]. Differences were positive in 13/14 and 16/16 people. The fixed shared rule also required positive association gain and class skill in each modality. It passed, p=0.00378418. Eight adults from another MRI study gave a 6.586-point mean difference on the same MRI sounds. Initial unshrunk development failed; its outcomes and the shrinkage amendment remain. An initial helper computed uninspected site losses before subsetting development people. No reserved person result informed choices. This is an internal endpoint holdout in reused studies. Electrical early-window and single-site checks were uncertain. Without human vocal sounds, the electrical class-gain interval included zero. Pre-event subtraction retained a positive regional mean interval. An exact error model gave different normalized regional gains with identical true tuning. Thus this is a restricted prediction result, not proof of a categorical neural mechanism or common cause. Sparse clinical coverage, source selection, separate libraries and cohorts, and unequal measurement filters limit scope. Code, permitted inputs, source pins and terms, all outcomes, figures and repeat steps are supplied.
human auditory cortex · intracranial EEG · fMRI · sound category · prediction · calibration · cross-modal generalization · measurement noise
References
16- usesdoi:10.1016/j.cub.2026.07.063
Human intracranial source study.
- usesdoi:10.5281/zenodo.20599001
Version 1.1 public CC BY electrical derivatives and source metadata.
- usesdoi:10.1152/jn.00588.2020
Primary B2021 fMRI study.
- usesdoi:10.1016/j.neuron.2015.11.035
Additional NH2015 fMRI study and MRI sound corpus.
- usesdoi:10.1371/journal.pbio.3002366
Pinned author neural archive, anatomy metadata and source processing.
- usesdoi:10.1016/j.cub.2022.01.069
Pinned MRI class and stimulus-ID reference.
- doi:10.1073/pnas.2412243122
Prior evidence for an auditory processing hierarchy; not a new confirmation.
- usesUp to four class examples reduced mean response pattern error after two amplitude controls in 20 iEEG patients and 16 fMRI adults sci:sha256:7022157afe1bb9ba7eea43f7a092273baa1257440813810284a25982d5525667
Prior class-example assay and fixed 0.25 shrinkage motivate the new regional test. Prediction procedures differ.
- Combined sound features gave positive mean gains over both single sets in 11 iEEG patients and eight fMRI adults sci:sha256:d3f382a918d33ee6d84d34c4e83b6ac1b02dc9350f35af96200eb63c734ddb9a
Related combined acoustic and category prediction results.
- usesMaps fit to other sound classes predicted fixed animal sound patterns in 20 iEEG patients and 16 fMRI adults sci:sha256:1b8693e8389d334be8780044f5056e91081fd0c71bf35ea03565a0b417773b48
Verified licensed electrical means, labels, baseline, MRI acoustic table, and MIT source-audit and extraction helpers.
- usesOne bit per auditory site gave mean four-class accuracy above chance in 20 iEEG patients and 16 fMRI adults sci:sha256:640399d48b2742bcc4d3c828910fda08a928b409a54c25a630f3a8b53642c0c2
Original electrical extract donor and exact source chunk map, via cf4efd8f.
- usesTwo fixed acoustic rules did not confirm shared nonvoice prediction gains in 18 SEEG patients and two fMRI groups sci:sha256:f6ade96a55abd5dd6909603449b3d63ad197279407aa009d924a3be21ddc98ae
Original licensed MRI eGeMAPSv02 measurements, via cf4efd8f.
- doi:10.1038/nature18933
HCP cortical parcel definitions and anatomical interpretation.
- usesdoi:10.1109/TAFFC.2015.2457417
Acoustic descriptor family.
- usesdoi:10.1145/1873951.1874246
Software used for the inherited MRI descriptor measurement.
- doi:10.7554/eLife.82566
Distinct person and stimulus uncertainty.
Cited by
7 publicationsPublic catalog entries on Sciscape.
- usesSound labels lowered error for omitted acoustic clusters in 18 iEEG patients and two fMRI groups
Licensed compact electrical data, acoustic measurements, source crosswalks and source-reader structure; prior random-calibration class test.
- usesAt 20 calibration sounds, matched class examples lowered association forecast error in 14 iEEG patients and 16 fMRI adults
Licensed compact inputs, reader and regional aggregation code; our new test jointly matches calibration examples and compares class coverage within a forecast rule.
- usesA fixed sound group forecast failed a shared test in 14 SEEG patients and 16 fMRI adults
We use its verified licensed electrical extract, source crosswalks, preparation code and prior development rule. We test a different four-group pooled-reference comparison.
- 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
Licensed compact data, source readers, source pins and class crosswalks.
- usesA fixed test did not confirm a shared sound class effect in 14 iEEG patients and 16 fMRI adults
We reuse its licensed compact inputs, MRI preparation code, label crosswalks and source audit code. We test a different endpoint.
- usesVocal labels failed shared forecast tests in 14 iEEG patients and 16 fMRI adults
We reuse its licensed compact inputs, source maps, regional task and forecast budget. We test binary label compression and exact equal-count risk.
- For fixed sounds, mean acoustic transfer gains were positive in SEEG and two fMRI studies; the test for every class failed
Earlier regional class-example prediction result; informs interpretation of the different acoustic-detail task.