BioSpine 2.0: A multimodal non-invasive neurorehabilitation program for people with chronic spinal cord injury

Study Devices The following non-invasive devices will be used as part of the intervention: Wearable wireless EEG headset (custom-developed at Griffith University, with dry electrodes; or equivalent commercial device). This device non-invasively records low-voltage biosignals from the surface of the scalp using electrodes that are compatible with hair and optimised for comfort. Functional electrical stimulation (FES) uses an electric current to externally stimulate muscles to facilitate movement. Surface electrodes consisting of adhesive hypoallergenic patches (Pals Platinum, Axelgaard Manufacturing Co Ltd, USA) are applied over the muscle groups of the lower limbs bilaterally (quadriceps, hamstrings, calves, and tibialis anterior). The electrodes do not cause damage to the skin or skin surface. FES is non-invasive and widely used in rehabilitation. The RehaStim 2 or P24 Science (Hasomed, Magdeburg, Germany) will be used for this intervention. Transcutaneous spinal cord stimulation uses an electric current delivered via self-adhesive surface electrodes (ValuTrode, Axelgaard Manufacturing Co Ltd, USA) placed on the dorsal aspect of the back overlying the thoracic (T11–T12) and lumbar (L1–L2) vertebrae, with return electrodes placed bilaterally over the iliac crests. Stimulation frequency ranges from 5–40 Hz and amplitude from 20–130 mA, individually optimised for each participant. This device is non-invasive and does not cause damage to the skin or underlying tissues. Arm and leg cycling ergometer is a stationary motorised bicycle providing pedal and handle assistance. The RT300 (Restorative Therapies, Baltimore, US), THERA-Trainer Tigo (THERA-Trainer, Hochdorf, Germany), or equivalent device will be used. The device is wheelchair compatible and enables participants to perform combined upper and lower limb cycling without requiring a transfer. Virtual reality (VR) involves visual and auditory simulated content delivered via a head-mounted display (Meta Quest 2/3, Facebook Technologies, LLC, or equivalent). VR provides a first-person view of the participant's avatar immersed in a naturalistic environment, creating an engaging and immersive cycling experience during rehabilitation. BioWrap® is a neoprene garment developed at Griffith University to facilitate rapid and reproducible placement of stimulation electrodes and sensors, improve comfort, and reduce session setup time. Dosage The study comprises four consecutive 6-month interventions, for a total study duration of up to 24 months. Participants may enrol at the beginning of any intervention and participate in consecutive or non-consecutive blocks of 6-month interventions, independently of whether they belong to the drug arm or the control arm of the intervention. Within each 6-month intervention, participants will perform at least 150 minutes per week of progressive neurorehabilitation, distributed over 2 to 3 sessions per week. Each session lasts between 60 and 75 minutes. Participants enrolled in the drug arm of the study will be administered 7.5 mg of buspirone hydrochloride orally twice daily, commencing at least one week prior to the start of the intervention and continuing throughout. Medication adherence will be monitored at each in-person session by querying participants about tablet intake since the previous session, and by requesting the return of unused tablets at the end of each month. Participants in the non-pharmacological arm will receive the same device-based intervention without buspirone. The allocation to the drug arm or control arm of the study is determined by participant's choice in consultation with their primary physician. Participants elected to participate in one arm of the study cannot enroll in a different arm of the study if re-enrolling in one of the subsequent 6-month intervention blocks, but can re-enroll in the same arm. Intervention At least two weeks prior to commencing each intervention, participants will undergo an individualised optimisation protocol for electrical stimulation parameters. During 1–2 weeks of visits (2–3 sessions per week, approximately 1 hour each), spinal transcutaneous stimulation will be delivered in supine and seated positions while surface electromyography records spinally evoked potentials from key lower limb muscles. Stimulation amplitude and frequency combinations will be systematically assessed to identify the optimal recruitment profile for each participant's sensorimotor pools below the level of injury. These personalised parameters will be used throughout the intervention and reassessed monthly. Each rehabilitation session is conducted by an experienced exercise physiologist or physiotherapist. At the beginning of each session, the participant is instrumented with FES surface electrodes on the main muscles of the lower limbs bilaterally, and self-adhesive spinal stimulation electrodes on the thoracic and lumbar spine. Wireless EMG and motion capture sensors may also be placed on the lower limbs to assess performance. A BioWrap® garment is used to securely hold electrodes and sensors in place. The participant approaches the cycling ergometer in their wheelchair (provided if necessary) and is secured to the device. They are then donned with the EEG headset and VR headset. The session begins with a passive cycling warmup of at least 5 minutes to reduce spasticity, during which no electrical stimulation is applied and motorised assistance is set to its minimum. Cadence begins at 20 revolutions per minute and is progressively increased toward a target maximum of 50 revolutions per minute (or a lower individualised maximum). The warmup concludes once the minimum duration has elapsed, the target cadence is reached, and no spasms are detected for at least 1 minute. A brain-computer interface (BCI) calibration procedure (varying in length between 1 and 25 minutes) is then performed within the session wherein a virtual avatar performing cycling or rest is displayed, and the participant is asked to imagine performing the visualised movement. Following calibration, the training phase commences: the participant imagines cycling, and this motor intent — as decoded by the BCI — triggers coordinated electrical stimulation of the lower limb muscles and the spinal cord, while the motorised ergometer simultaneously provides pedal and handle assistance. BCI calibration data may be automatically updated during the training phase. Ergometer torque is continuously monitored to detect spasms; upon detection, electrical stimulation and motorised assistance are gracefully shut down to prevent the application of force during a spastic event. Parameters of Electrical Stimulation FES electrodes are positioned bilaterally on the hamstrings, quadriceps, calves, and tibialis anterior. Stimulation frequency is set to 25 Hz, with amplitude individually identified by manual muscle palpation for twitch response and capped at 80 mA. The timing and amplitude of stimulation is automatically modulated throughout the pedalling cycle to activate each muscle of the right amount to achieve the target cycling cadence only when muscles are in their mechanically advantageous phase. For transcutaneous spinal cord stimulation, cathodal electrodes are placed over T11–L2 vertebrae and anodal return electrodes over both iliac crests. Stimulation frequency ranges from 5–40 Hz and amplitude from 20–130 mA, based on individually optimised parameters. Spinal stimulation is delivered synchronously with BCI-decoded motor intent and thereby delivered only when the participant is imagining to cycle..