Lateralized excitation-inhibition rebalance correlates with motor recovery following hemispheric surgery

Hemispheric epilepsy surgery effectively achieves seizure control, yet post-operative motor recovery remains challenging. Whether motor compensation is linked to the excitation-inhibition balance in the unaffected hemisphere (UH) remains unclear. Using scalp EEG, we quantified the aperiodic exponent of the power spectrum as a non-invasive biomarker of cortical excitation-inhibition ratio (where a flatter spectral slope reflects neural excitation). We retrospectively analysed 46 patients who underwent hemispheric surgery and 23 age-matched unilateral non-hemispheric controls. Motor function was measured preoperatively and 3 months post-operatively, with preoperative status classified as preserved or impaired. We assessed preoperative inter-hemispheric asymmetry, post-operative UH modulation and their linear interaction in predicting motor improvement. Jackknife mapping was used to identify the contribution of a specific scalp sensor. Our results show that the hemispheric cohort exhibited a distinct preoperative excitation-inhibition asymmetry (P < 0.0001). Post-operatively, the UH exponent interacted significantly with preoperative status to track the change of motor function (P < 0.0001; R ² = 0.53). In patients with preoperative deficits, flatter UH slopes (disinhibition) were associated with functional gains (P = 0.0442), while steeper slopes (over-inhibition) correlated with decline (P = 0.0001). Conversely, in patients with preserved function, steeper UH slopes (inhibition) were linked to better outcomes (P = 0.0023), whereas flatter slopes led to deterioration (P = 0.0175). Jackknife mapping localized the primary contribution to the sensor overlying the UH centro-parietal region. These findings suggest that state-dependent modulation of UH excitation-inhibition balance is a factor associated with motor recovery after hemispheric surgery-stabilization is beneficial when baseline function is preserved, while facilitation is beneficial when it is impaired. The aperiodic exponent emerges as a practical biomarker to monitor cortical excitation-inhibition dynamics and to guide individualized, precision neurorehabilitation strategies.