Post-implantation 3-T MRI in patients with an epicranial stimulation device: image quality and deep learning-based electrode reconstruction

We evaluated structural and functional MRI at 3 T in five patients with an implanted EASEE focal cortex stimulation system and assessed MRI-only electrode reconstruction. Image quality was evaluated visually and by comparing implant-centered and control regions using structural GM/WM signal ratios, GE-EPI signal intensity, T2*, and relative off-resonance. A deep learning pipeline trained exclusively on synthetic data estimated electrode position from T1-weighted MPRAGE images. Structural and functional acquisitions were completed in all patients without scan-related adverse events. Structural images showed clear intracranial anatomy beneath the implant without visually apparent local signal loss or distortion. GE-EPI reference images and derived maps showed no consistent local disturbance corresponding to the electrode. GE-EPI signal was numerically lower in implant-centered regions, but no imaging metric differed significantly between regions (all p > 0.05). Electrode reconstructions were anatomically plausible in all patients, with patient-specific median central electrode-to-cortical gray-matter distances ranging from 9 to 20 mm. These findings demonstrate the feasibility of post-implantation structural and functional MRI and MRI-only estimation of electrode-cortex geometry, although localization accuracy remains unvalidated. This combination provides a basis for future studies integrating patient-specific electrode geometry, electric-field modeling, and functional connectivity analyses in focal cortex stimulation with the EASEE system.