Implant laterality did not predict five free-recall processes in 253 RAM SEEG patients

We tested whether an implant laterality index (LI), used as a proxy for suspected seizure side, predicted temporal contiguity (TF), semantic clustering (SEM; categorized free recall only), primacy, recency or intrusions in session 0 of RAM SEEG patients. We used event files from OpenNeuro ds004789 (FR1) and ds004809 (catFR1), plus the patient LI table from a prior internal study. The analysis plan was written and hashed before scores were calculated. Partial Spearman tests adjusted for age, sex, dataset, site and log pair count; p values used 10,000 permutations and 95% confidence intervals used 5,000 patient bootstrap resamples. Among 253 included patients (221 with at least 10 valid TF transitions; 64 categorized free-recall patients for SEM), LI did not predict TF (n=221, rho=-0.106, 95% CI -0.240 to 0.025, p=0.118), with recall rate as a covariate (rho=-0.106, p=0.109), SEM (n=64, rho=0.017, p=0.897), primacy (rho=-0.032, p=0.606), recency (rho=0.005, p=0.938) or intrusions (rho=-0.001, p=0.990). Secondary Holm adjusted p values were at least 0.54. No exploratory test using |LI| had p<0.05; its smallest p was 0.062 for primacy. These results show no evidence that implant laterality predicts these recall processes, consistent with a general reduction in recall rather than a large, process-specific loss. The confidence intervals do not exclude small effects; SEM had only 64 patients and low power. LI is only a proxy for seizure side, and the data lack seizure-onset, language-dominance and lesion labels. These are between-patient associations from session 0. The included event files were downloaded from the public OpenNeuro S3 bucket on 2026-10-11; their source URLs and SHA-256 hashes are retained, but their exact dataset releases are uncertain.

free recall · intracranial EEG · stereo-electroencephalography · implant laterality · temporal contiguity · semantic clustering · primacy · recency · intrusions · RAM · OpenNeuro