Lateralized amyloid PET uptake and beta-band connectivity in relation to cognition in temporal lobe epilepsy.

We examined focus-relative amyloid PET asymmetry and its associations with resting-state EEG connectivity and processing speed in mesial temporal lobe epilepsy (MTLE). Thirty adults with clinically assigned MTLE (19 right-focus, 11 left-focus) underwent 18F-florapronol PET and resting-state EEG. Regional standardized uptake value ratios (SUVRs) and sensor-level weighted phase lag index (wPLI) were evaluated using paired comparisons, rank correlations and exploratory bootstrap indirect-association models. Cognitive data were available for 26 participants, including Digit Symbol Coding (DSC) for 25. Under the original analytical convention, mean medial temporal SUVR was 1.066 ipsilaterally and 1.013 contralaterally (difference 0.053, 95% CI [0.013, 0.093]; p = 0.011; Cohen's dz = 0.500). The contrast differed by focus side (interaction p = 0.002) and was 0.038 under source-table reconstruction (p = 0.075). Ipsilateral SUVR correlated with inter-hemispheric beta wPLI (ρ = 0.617, q = 0.034 across 240 tests). The intra-contralateral association (ρ = 0.536) did not survive this broader correction. Indirect estimates for Trail Making Test A (0.279, 95% bootstrap CI [0.043, 0.761]) and DSC (0.151, 95% bootstrap CI [-0.114, 0.476]) did not survive correction across 136 models. Regional medial temporal PET uptake was associated with beta-band scalp connectivity in this small cohort, while focus-side heterogeneity and labeling sensitivity limit a uniform ipsilateral interpretation. Processing-speed indirect associations were exploratory and did not establish compensation or causality. Replication with structural correction, longer EEG sampling and longitudinal cognitive follow-up is required.