In 21 RAM patients, implant laterality predicted verbal recall relative to spatial memory, but not spatial memory alone

In a pre-registered analysis of 28 patients with YC1 spatial-memory scores, implant laterality did not predict object-location memory (Spearman rho = +0.14, permutation p = 0.46, bootstrap 95% CI [-0.25, +0.52]). The test had 80% power only for effects larger than |rho| = 0.51, so a moderate spatial effect is not excluded. In the 21 patients with both YC1 and FR1 data, laterality predicted the standardized spatial-minus-verbal score (rho = +0.66, permutation p = 0.0016, Holm p = 0.0032, 95% CI [+0.27, +0.90]): more left-side depth contacts were associated with lower verbal recall relative to spatial memory. Secondary results linked laterality to verbal recall at rho = -0.41 (p = 0.071) and spatial memory at rho = -0.03 (p = 0.91); the spatial and verbal scores correlated at rho = +0.58. Thus, the interaction was driven by the verbal side, and the data do not show the spatial half of a double dissociation. Contact laterality is only a proxy for suspected seizure side because clinical seizure-onset labels are unavailable. The 21 patients are a subset of a prior verbal-memory sample, so this is not an independent replication. Small samples, ceiling and age effects in YC1, and the uncertainty of the exact OpenNeuro release limit interpretation.

intracranial EEG · memory · laterality · verbal recall · spatial memory · epilepsy