A sound profile test without a reference was uncertain in six SEEG patients and positive in 16 fMRI adults

We tested whether one common sound profile, with a separate gain for each record, can describe selected auditory population means. Separate human studies supplied SEEG high-gamma responses and BOLD fMRI estimates for 165 sounds. Two patients and four MRI adults supplied development. A local lock preceded new endpoint tests in six other patients and 16 other adults. A product of matching two-by-two response determinants avoids a fitted reference profile. Its expectation is zero for a single true profile under stated repeat-error assumptions. Mean W was 1.395103e-05, with approximate 95% person jackknife limits [-0.0001911453, 0.0002190473] in SEEG. It was 0.0003158737, [0.0001952998, 0.0004364475], in primary fMRI. The shared rule failed; Holm p=0.9675597. Two initial reference-based claims also failed. Eight adults from another MRI study had a positive descriptive W result. Electrical cleaning checks changed estimates greatly. All negative results remain. In one declared Gaussian two-profile model, the positive-limit rule passed in only 72 of 1000 six-person draws. A common multidimensional neural code with different site mixtures also gives positive W. Thus the test concerns scalar recorded profiles, not different personal neural codes or a shared neural cause. Small clinical samples, source repeat screens, unequal cohorts, spatial mixtures and different time scales limit inference. The package supplies the method, proof, code, source pins and terms, all results, figures and repeat steps. No positive cross-modal neural result was confirmed.

human auditory cortex · SEEG · fMRI · repeat measurements · population response · common profile · rank constraint · negative result · reproducible method