For fixed sounds, mean acoustic transfer gains were positive in SEEG and two fMRI studies; the test for every class failed

We tested whether an acoustic response model predicts differences among sounds in a class absent from neural training. Separate public studies supplied signed SEEG field means and BOLD fMRI. Four patients and four MRI adults supplied development. A local code lock preceded new scores in 21 patients, 16 B2021 adults and eight NH2015 adults. At ridge penalty 10, mean gains were 0.02807, 0.07546 and 0.07354 source within-class variance units. Their 95% person Student limits were [0.00952,0.04662], [0.06956,0.08136] and [0.06130,0.08579]. The fixed average rule also required an electrical advantage over a pre-event control. It passed two-claim Holm correction, p=0.004971. The fixed test for every class failed, p=0.917961. In 999 full sound refits with person resampling, the electrical range was [-0.00220,0.04917]. Thus the fixed-library person result did not confirm robust transfer over both new people and new sounds. An added regional comparison favored early over association parcels in all 38 people with both panels; it is exploratory. An added exact gain-curve tool found no confidence-supported common positive scale interval for all classes. Independent solvers rebuilt all 265 person-class fits, and original chunks rebuilt all 110682 electrical window values. Separate cohorts and libraries, clinical sampling, source screens, unequal filters and feature builds limit scope. No shared neural cause was established. The package supplies code, permitted inputs, source pins and terms, all results, figures and repeat steps.

auditory cortex · SEEG · fMRI · acoustic transfer · sound classes · prediction · negative result