A fixed hippocampal boundary response test failed in 12 iEEG patients and ten fMRI adults

We tested whether hippocampal activity rises at annotated film boundaries after camera cuts and adjacent shot duration are controlled. Separate public human studies supplied bipolar intracranial EEG power during an eight-minute film and BOLD fMRI during a two-hour film. Development used four patients and five adults. A fixed new-endpoint test used 12 other patients and ten other adults. Mean adjusted pre-to-post rises were +0.08838 and +0.08334. Their 95% person t limits were -0.02199 to +0.19875 and +0.03388 to +0.13279. Scores were positive in 8/12 and 9/10 people. The shared mean/sign p was 0.193848. Both person/block stability ranges included zero. The shared rule failed. A later design audit found a separate method limit. Under fixed positive kernel shapes, the fitted common rise was a signed mean of event amplitudes. Four of twelve electrical event weights and sixteen of 162 MRI event weights were negative. All-positive event rises with amplitude ranges 1–8 and 1–34 gave fitted common rises of -0.25335 and -0.06359. Independent dense fits and an optimization check confirmed the construction. Exact sign-change bounds and all kernel checks are supplied. These are model examples, not measured human event amplitudes or a new regression theorem. Unequal boundary labels, sparse electrical events, clinical sampling, separate films and measurement filters limit scope. No positive cross-modal neural or causal claim is made. The package gives code, permitted compact electrical inputs, new MRI model summaries, source pins and terms, all scores, figures, local plans and repeat steps. Original MRI signals remain external.

hippocampus · event boundaries · intracranial EEG · fMRI · response overlap · event weights · reproducibility