Differential Functional Network Remodeling Induced by Low- and High-Frequency rTMS: Evidence From Concurrent fNIRS Monitoring

Repetitive transcranial magnetic stimulation (rTMS) at low-frequency (LF) and high-frequency (HF) has been shown to facilitate motor recovery after stroke, however the underlying neural network mechanism remains unclear. This study employed functional near-infrared spectroscopy (fNIRS) to monitor hemodynamic changes in real-time during rTMS, aiming to evaluate the immediate effects of LF- and HF-rTMS on functional network remodeling, and to explore the long-term impact of rTMS-induced neural changes on motor function recovery. A total of 108 stroke patients were randomly assigned to LF-rTMS, HF-rTMS or Sham groups and received 15-days of rTMS intervention. fNIRS was employed to detect hemodynamic changes during the intervention. The laterality index (LI) and the wavelet phase coherence (WPCO), based on wavelet transform, were used to describe functional network remodeling. Clinical scales were used to evaluate patients' behavioral outcomes. LF-rTMS significantly increased LI during the first intervention and induced WPCO changes between motor regions. In contrast, HF-rTMS produced delayed yet significant alterations in WPCO, with long-term intervention modulating both motor and cognitive networks. After 15 days, both LF- and HF-rTMS showed significant behavioral improvements correlated with WPCO changes. The rTMS-fNIRS approach provided neural mechanistic evidence for the role of rTMS in promoting functional recovery. LF-rTMS mitigates abnormal interhemispheric inhibition and induces behavioral improvements during the short-term treatment process. In contrast, HF-rTMS requires sustained stimulation to achieve remodeling effects but may offer broader rehabilitative benefits. These behavioral changes result from acute neural modulation induced by rTMS, which may consolidate transient plasticity into long-term motor recovery through repeated interventions.