After one heat score was removed, fixed pain readers failed a shared test in 119 EEG and 20 fMRI adults

We tested whether a second neural score tracks pain reports after control of stimulus strength, context, order and a learned neural heat score. Separate public studies supplied scalp EEG and fMRI trial estimates. Development used 39 EEG and eight MRI adults. A local code lock preceded tests in 119 other EEG and 20 other MRI adults. Mean partial correlations were 0.04294 and 0.08561. Their pointwise 95% person Student intervals were [0.00351,0.08237] and [0.03202,0.13919]. Mean error reductions were 0.102 and 0.572 points of adjusted-report variance; both intervals crossed zero. The full shared rule failed, p=0.28687. Whole-person resampling with both readers refitted gave correlation ranges across zero in both modalities. A fixed EEG reader at visit two had uncertain mean correlation. A separate EEG study with 106 eligible adults also failed its full corrected rule, Holm p=0.06360. All negative people and controls remain. Exact score and noisy-measurement examples state limits on interpretation. Independent calculations rebuilt all 139 primary person results within 5e-15. The package supplies compact permitted inputs, source pins and terms, code, fixed weights, plans, results, figures and repeat steps. These are offline scores that use target report residual scales, not absolute pain forecasts. No positive cross-modal prediction result, distinct pain process or causal mechanism was confirmed. Cohort differences, source selection, measurement filters, coarse MRI sampling and incomplete nuisance records limit scope.

pain · EEG · fMRI · heat · residual prediction · measurement error · negative result · reproducibility