Hemiparetic cerebral palsy is a condition in which one side of a child's body becomes weak due to brain injury occurring around birth, often caused by stroke. Weakness on the affected side reduces the ability to use the involved hand and arm, limiting everyday activities such as dressing, eating, and playing. Few effective treatments currently exist, especially for children with severe weakness.

The study described here examines a new therapy that combines Brain-Computer Interface (BCI) technology with Functional Electrical Stimulation (FES). BCI technology uses brain signals to control external devices; in this therapy, brain signals activate muscles through electrical stimulation. When a child imagines wrist movement, the system detects the associated brain activity and delivers electrical stimulation to generate actual movement. Such paired activity supports neural rewiring and strengthens connections between the brain and muscles, leading to improved arm function.

Previous research demonstrates strong benefits of BCI-FES for adults after stroke, but minimal testing has been conducted with children. The current study will evaluate whether BCI-FES improves arm and hand function in children aged 12 to 17 with hemiparetic cerebral palsy.

Participants will complete 15 to 20 sessions over a two-month period while wearing a cap that records brain signals. During each session, the system provides muscle stimulation and visual feedback through animated hand movements. Outcome measures will include performance of daily tasks, hand dexterity, muscle activity, and the presence of any adverse effects.

The overarching goal is to create a fun, engaging, and effective therapy that supports recovery of hand use and greater independence. Successful results could guide larger studies and inspire new technology-based treatments that enhance quality of life for children with cerebral palsy.

Purpose Hemiparetic cerebral palsy leads to severe unilateral motor weakness that impairs upper limb function and daily living in children. Effective, evidence-based therapies-particularly engaging, intensive interventions such as Brain-Computer Interface combined with Functional Electrical Stimulation (BCI-FES)-remain limited and underexplored in pediatric populations. The study aims to address this gap by adapting and validating a child-centered BCI-FES rehabilitation model for upper-extremity motor impairments, enabling effective and engaging therapy for children with hemiparetic cerebral palsy.

Objectives Aim 1: Determine the efficacy of BCI-activated FES for children with hemiparetic cerebral palsy in improving functional outcomes.

Aim 2: Develop, validate, and optimize a child-centered BCI-FES rehabilitation model that improves upper-extremity motor function and daily living activities in children with hemiparetic cerebral palsy, enabling scalable, effective neurotechnology-based therapies that enhance functional independence and quality of life within pediatric rehabilitation.

Method The setup combines EEG, FES, and real-time visual feedback to create a closed-loop training system. During each session, an EEG cap records motor-related brain activity while a participant imagines a specific movement (for example, wrist extension). When the system correctly identifies motor imagery, the system delivers FES to the corresponding muscles and displays the associated movement on a screen through an animated avatar.

The project uses the g.tec recoveriX system (g.tec, Schiedlberg, Austria), a commercially available BCI suite that integrates EEG recording and functional electrical stimulation.

BCI: EEG signals are recorded using a 16-channel, gel-based EEG headset. Recording electrodes are placed according to the international 10-10 system. The classification algorithm generates classification accuracy, defined as the percentage of correct predictions produced by the BCI when identifying intended motor-imagery commands.

FES: Electrical muscle stimulation is delivered bilaterally through two g.Estim electrical stimulators (g.tec, Graz, Austria), which generate rectangular biphasic constant-current pulses.

Visual Feedback: A monitor positioned directly in front of each participant displays two virtual hand avatars that provide real-time visual feedback mirroring intended movement. When the system successfully detects motor imagery, the avatar performs the corresponding movement on the screen.

Intervention The intervention consists of two components: Training Sessions and Rehabilitation Sessions.

Training Sessions: Training calibrates the system by recording EEG signals during motor imagery of left or right wrist extension. Participants are guided to visualize wrist extension for both hands. Each session includes three trials of approximately 15 minutes. Each trial contains 80 runs lasting 8 seconds.

Rehabilitation Sessions: During rehabilitation, the classifier generated during training drives visual and electrical feedback. Participants complete up to three trials per session, using audio cues to guide motor-imagery attempts. Visual feedback and electrical stimulation are provided only when motor imagery is accurately classified and matches the cue direction.

Participants attend 15 to 20 recoveriX sessions over three consecutive weeks. Each session lasts up to 90 minutes and includes randomized motor-imagery trials involving both limbs. Sessions are adjusted according to participant tolerance, with scheduled rest periods and optional additional breaks to maintain comfort and engagement.

Inclusion Criteria:

1. Clinically confirmed and functionally disabling moderate to severe hemiparetic CP (GMFCS levels I-IV and MACS levels II&III) with personalized goals that include improvement of wrist extension of affected upper extremity
2. Age 12-17 years,
3. Informed Consent/assent,
4. Normal corrected vision and hearing
5. Ability to maintain supported sitting for 30 min or more

Exclusion Criteria:

1. Bilateral perinatal stroke
2. Motor strength of wrist extension zero
3. Severe hemiparesis (MACS V)
4. Severe developmental delay and/or other inability to comply with study protocol
5. Severe wrist contractures limiting wrist extension
6. Upper extremity orthopedic surgery or botulinum toxin in the preceding 6 months,
7. Initial classification accuracy below level of significance (as determined by the BCI system)