Peak width of skeletonized mean diffusivity as a quantitative imaging marker of white matter alterations in MRI-negative temporal lobe epilepsy: validation with demographic covariate adjustment

This study evaluated the peak width of skeletonized mean diffusivity (PSMD) as a quantitative imaging marker in patients with magnetic resonance imaging-negative (MRI-negative) temporal lobe epilepsy (TLE), accounting for age and sex effects.

This retrospective study included 34 MRI-negative TLE patients and 40 age- and sex-matched healthy controls (HCs). PSMD was calculated using tract-based spatial statistics and histogram analysis. Hierarchical regression identified PSMD determinants. Receiver operating characteristic analysis compared uncorrected and age- and sex-corrected PSMD, with the latter validated by 10-fold cross-validation.

PSMD was significantly elevated in patients versus HCs (2.545 × 10^-⁴ mm²/s vs. 2.340 × 10^-⁴ mm²/s, p = 0.004) and independently associated with TLE (OR = 2.39, p = 0.011). Sensitivity analysis confirmed robustness (OR = 2.22, p = 0.013). In addition, males showed higher PSMD than females in both patients (p = 0.022) and HCs (p < 0.001). Furthermore, age (β = 0.453, p = 0.005) and male sex (β = 0.539, p = 0.014) were associated with higher PSMD. Age/sex correction improved specificity (82.5% vs. 67.5%) but reduced sensitivity (64.7% vs. 52.9%), with an area under the curve (AUC) of 0.696. AUC improvement was non-significant (p = 0.676); cross-validated AUC was 0.628.

PSMD is a quantitative marker of diffuse white matter pathology in MRI-negative TLE, modulated by demographic factors. Consistent sex effects across groups indicate biological rather than disease-specific variation. Demographic adjustment improves specificity but not overall accuracy, supporting PSMD for group-level cross-sectional studies. Sex effects must be controlled for in white-matter diffusion MRI.