Calibrated extra axes reduced mean repeat error in 16-site panels from six SEEG patients and 24 fMRI adults

We tested whether added auditory response axes improve repeat prediction in separate human studies. Training sounds set site axes. Separate calibration sounds set a weight for axes five through eight. Other sounds tested prediction from one repeat to another. Twenty fixed 16-site panels gave one score per person. Development used two SEEG patients and four fMRI adults. A local code lock preceded tests in six other patients and 24 adults from two fMRI studies. Mean gains over four fixed axes were 0.3776%, 0.4195% and 0.3278% of measured target energy. Their 95% person Student intervals were [0.1192%,0.6360%], [0.3205%,0.5184%] and [0.2068%,0.4488%]. All 30 people also improved against matched fixed eight-axis readers. The shared claim passed two-claim Holm correction, p=0.03125. The original fixed-weight claim failed; all 30 scores were negative. The short pre-onset SEEG control was negative against four axes. Artificial same-data selection gave gains with independent noise. A rank-one signal example also gave an extra-axis gain. Thus the result concerns selected processed measurements, not neural rank or a common cause. Separate cohorts, source screens, unequal sampling and different time filters limit scope. Code, exact source pins and terms, plans, all results, figures and repeat instructions are supplied. Original neural arrays stay external.

human auditory cortex · SEEG · fMRI · repeat prediction · cross-modal test · component shrinkage · source selection