Low-intensity focused ultrasound versus rTMS for motor recovery in subacute stroke

BACKGROUND: Low-intensity focused ultrasound (LIFU) is a non-invasive neuromodulation technique with high spatial precision and the ability to reach deeper brain regions, offering potential advantages for post-stroke rehabilitation. Repetitive transcranial magnetic stimulation (rTMS) is a widely adopted non-invasive brain stimulation technique that modulates cortical excitability to promote neuroplasticity. However, direct head-to-head comparisons between these two modalities for post-stroke motor recovery remain limited. OBJECTIVE: To perform a secondary head-to-head comparison of LIFU and repetitive transcranial magnetic stimulation (rTMS) for motor recovery after stroke, based on a prospectively registered randomized controlled trial. METHODS: This secondary analysis included patients with subacute stroke who received two weeks of standard rehabilitation combined with either LIFU (n = 25) or rTMS (n = 25) targeting the ipsilesional primary motor cortex. LIFU parameters: 0.5 MHz, spatial-peak pulse-average intensity (ISPPA) 10.2 W/cm² (free-field), pulse duration 0.2 ms, duty cycle 20%, 20 minutes per session, five days per week for two weeks (10 sessions total). rTMS parameters: 10 Hz, 80% resting motor threshold, 1,000 pulses per session (20 trains of 5 seconds), 20 minutes per session, five days per week for two weeks (10 sessions total). Motor outcomes were assessed using the Fugl-Meyer Assessment (FMA; upper and lower extremities), Modified Barthel Index (MBI), and Brunnstrom stages. Resting-state functional near-infrared spectroscopy (fNIRS) was used to evaluate cortical activity and functional connectivity before and after the intervention. Primary analyses were conducted in the intention-to-treat (ITT) population (n = 50), with completer analyses (n = 43) performed as sensitivity analyses. RESULTS: Both groups showed significant within-group improvements in FMA and MBI after the intervention (all p < 0.001), and changes in Brunnstrom stages were also uniformly significant. No statistically significant between-group differences were observed in post-intervention FMA, MBI, or Brunnstrom stages (all p > 0.05), and completer analyses yielded consistent between-group conclusions. In contrast, change-from-baseline analyses demonstrated greater improvements in FMA scores in the LIFU group compared with the rTMS group (ΔFMA upper limb: median 7 [IQR 3-10.5] vs. 2 [1-3], p = 0.001; lower limb: 3 [1-4.5] vs. 1 [0-1.5], p < 0.001). Exploratory fNIRS analyses revealed modality-specific patterns: prefrontal fractional amplitude of low-frequency fluctuations (fALFF) increased significantly in the LIFU group (p = 0.002) but not in the rTMS group, while functional connectivity changes did not remain significant after correction for multiple comparisons. CONCLUSION: LIFU and rTMS were associated with comparable short-term motor outcomes in subacute stroke. Differences observed in change-from-baseline motor improvements and exploratory neuroimaging measures suggest potential divergence in recovery dynamics and cortical modulation, warranting further investigation in larger, longitudinal studies. TRIAL REGISTRATION: This study was derived from a prospectively registered, three-arm randomized controlled trial in the Chinese Clinical Trial Registry (ChiCTR2500114687). The present manuscript reports a secondary head-to-head comparison between the two neuromodulation intervention arms.