A 65-year-old man had a history of hypertension for 4 years. He did not take any medication or monitor his blood pressure regularly. He was admitted to the First People’s Hospital of Kunming, China due to sudden slurred speech, nausea, and vomiting. Upon physical examination, the patient was unresponsive with the Glasgow Coma Scale (GCS) of E1 (no eye-opening), V1 (no verbal response), and M1 (no motor response) to any kind of stimuli. Bilateral pupils were dilated, and pupillary light reflex was present. Muscle strength in the extremities could not be measured. His clinical assessment with the NIH Stroke Scale (NIHSS) score was 32, implicating a severe stroke. A computed tomography (CT) scan of the head showed massive intracerebral hemorrhage in the left frontal, temporal, and parietal lobes with cerebral herniation. The diagnosis was massive cerebral hemorrhage with cerebral herniation secondary to high-risk, grade 3 hypertension. He was subjected to decompressive craniectomy and hematoma removal. After surgery, he received 2 weeks of treatments including a tracheotomy, medications for anti-infection, anti-hypertensive, and neurotrophic drugs, and hyperosmolar therapy with mannitol, and his GCS was improved to E3 (opens eyes in response to voice), VT (under tracheotomy), and M5 (move to localize pain). However, the patient experienced cognitive, speech, and swallowing disorders, irritability, agitation, behavioral disorder, urinary and bowel incontinence, and right limb immobility. The CT scan showed irregular cerebral edema in the left frontal lobe, temporal lobe, parietal lobe, and posterior limb of the internal capsule, and mild meningoencephalocele after left frontotemporal parietal bone flap removal (,). The patient was sent to the Second People’s Hospital of Kunming, China for rehabilitation treatment. He received conventional rehabilitation therapy including cognitive, swallowing, speech, physical therapy, occupational therapy, median nerve surface electrical stimulation, and traditional acupuncture therapy (total treatment time of 4 h per session, once session a day, 5 times a week). After continued active rehabilitation treatment for 3 months, the patient’s cognitive, speech, and swallowing were significantly improved. He could complete one-step instructions, give simple answers, and able to tolerate orally well. The Brunnstrom Motor Staging (BMS) of the upper limb, hand, and lower extremities of the right limb was at stage I-I-IV. The muscle strength of his right lower limb was at grade 3, and he could walk with minimal support and assistance. However, the muscle strength of his right upper limb was at grade 0 and there was no autonomous activity. The patient’s Modified Barthel Index (MBI) score was 31, Fugl-Meyer Assessment for upper extremity motor function (FMA-UE) score was 4, and he was unable to perform any of the tasks in Jebsen Hand Function Test (JHFT) (). After explaining the purpose of the treatment, the procedure, and the possible risks to the patient and his family, and obtaining their consent, the patient received ultrasound-guided right median nerve electrical stimulation (UG-MNES). Upon receiving the UG-MNES treatment, the patient lay on his back, the medial side of the right forearm was fully exposed, and 7–10 cm above the Rascette lines was scanned repeatedly with short axial ultrasound to locate the median nerve. The median nerve is a circular honeycomb structure between the flexor digitorum superficialis and the flexor digitorum profundus. The median nerve of the affected upper limb was probed with a high-frequency ultrasound. The probe was adjusted until a clear image of the median nerve was observed. Electrical stimulation was delivered through a peripheral nerve stimulator (SY-708A, Su Yun, Jiangsu, China), which was pre-loaded with a disposable introducer needle (size 0.5 mm × 50 mm). The needle was inserted in-plane, avoiding the blood vessels and tendons, until the needle tip was attached to the nerve sheath as seen on ultrasound, the needle core was withdrawn, and the peripheral nerve stimulator was connected to deliver a bidirectional rectangular wave with a stimulation frequency of 2 Hz and wave width of 0.2 ms for 20 min. The stimulus current was adjusted to 1.0 mA for 5 min, followed by 1.5 mA for 15 min to trigger the thumb and forefinger palm flexion movement. This intervention was performed once a week for a total of four sessions (,). During this 4-week intervention, the patient received conventional rehabilitation treatments except for median nerve surface electrical stimulation which was replaced by UG-MNES. Similar to the assessment conducted before treatment (time-point i), the BMS, JHFT, FMA-UE, and MBI were used to assess the functional recovery of the right upper limb and hand after treatment at three different time-point of measurement, i.e., immediately after the first intervention to determine the effectiveness of UG-MNES treatment (time-point ii), 1 week after the second intervention (time-point iii), and 1 week after the fourth intervention (time-point iv). The third and fourth time-point measurements were used to determine the long-term functional recovery after UG-MNES treatment. Electromyography (EMG) and nerve conduction velocity (NCV) tests were performed to assess nerve damage and dysfunction before treatment and after the complete course of treatment (). Immediately after the first intervention, the BMS of the right upper limb and hand had improved from I-I (before treatment) to IV-IV. The patient was unable to perform any of the tasks in JHFT before treatment and regained the ability to perform all the tasks in JHFT, except moving heavy objects after the first intervention. The MBI score improved from 31 (before treatment, severe dependence) to 50, and the patient had made great progress but required assistance in feeding, toilet transfers, grooming, bathing, and dressing. The FMA score had also improved from 4 (before treatment) to 23, and immediate flexion and extension of the upper limb were observed (;,). The long-term functional recovery after UG-MNES treatment was assessed 1 week after the second and fourth interventions. The results showed that 2-week of rehabilitation was insufficient to facilitate long-term functional improvement and the motor function recovery effect was observed after 4-week of rehabilitation. After 2-week of rehabilitation, although the patient was able to perform all the tasks in JHFT, his BMS was III-III, his MBI score was 46, and his FMA score was 18. At the end of the 4 weeks of rehabilitation, his BMS was V-IV and the time to perform the tasks in JHFT was reduced. The MBI score had increased to 85 (moderate dependence) and the patient can personally perform self-feeding and ambulation, and need minimal supervision in chair/bed transfer, toilet transfers, grooming, bathing, stairs climbing, dressing, bladder control, and bowel control. The FMA score was 61, suggesting improvement in motor functioning, stability, balance, and joint functioning in the upper extremity, wrist, and hand (). The EMG of abductor pollicis brevis (APB) muscle innervated by the median nerve showed normal spontaneous activity and interference pattern before and after UG-MNES treatment. Furthermore, based on the motor nerve conduction velocity (MNCV) and sensory nerve conduction velocity (SNCV), the onset latency and amplitude of compound muscle action potential (CMAP) and sensory nerve action potential (SNAP) before and after treatment were normal. The results of EMG and NCV indicated that UG-MNES did not cause nerve damage ().