A BCI-controlled soft actuated exoskeleton for tele-neurorehabilitation

Around 1 in 6 people live in the UK with a neurological condition, and an estimated 600,000 people are diagnosed with a neurological disorder each year [1]. Rehabilitation costs put a financial burden on the NHS. For instance, the total annual societal costs of supporting stroke patients worldwide are £26 billion per year, including £8.6 billion for the NHS and social care [2]. During the pandemic, the out-patient based face to face service was postponed for a substantial period to reduce the spread of COVID. A large number of patients or their carers became fearful of catching the virus and declined to attend face-to-face clinics when services resumed. Despite the neurological weakness, assessment and provision of rehabilitation exercises, posture management etc could have been attempted with these patients by video consultation. Qualitative and quantitative evidence of improvement in patients’ movement is necessary to make these interventions remotely. Following the clinical trial of rehabilitation by REX exoskeleton [3], it has been proven to be feasible and safe for people with SCI (spinal cord injury) and the acceptability of users is high. However, those existing motor-based heavy exoskeletons require licensed clinicians to guide and practice with the user, and unfortunately, there would not be more than such 4-5 experts in the UK. Also, the average weight of those exoskeletons is 50 kg, and the cost is more than £100K GBP. Soft actuator [4] based devices appear to be more suitable for rehabilitation purposes compared to the hardware-based motorised bulky device as it is a lightweight system, possesses a high torque to weight ratio and provides in-built compliance to joint movement. Devices such as Teslasuit [5] are only able to measure joint parameters from wearable sensors and provide electrical stimulation however this does not provide any physical assistance to the user.

Although a few SMA-based wearable devices [6] have been developed, there is no such affordable device delivering tele-neurorehabilitation to multiple degrees of freedom and also lacking BCI (Brain-computer interface) control in case of patients with severe brain injury. In terms of control, the proposed exoskeleton will use BCI signal [7]. The control mechanism will be composed of several modes of receiving the user's intention to trigger the system and fine control of the task trajectory depending on the condition of disability. For example, for people with severe disability where no muscle activation is possible, the exoskeleton can be triggered using only brain signals using EEG [8] while the low-level control can be generated by the corresponding muscle activations using EMG [9]. Currently, there is no online personalized setup available to guide and assist these patients through specific exercises in their homes along with monitoring and recording their performance.

The proposed project aims to develop a soft actuator-based wearable suit that would help patients to carry out rehabilitation services at home. Along with this, the device will be integrated with BCI (Brain-computer interface) functionality and tele-control features so that patients with severe brain injuries can take advantage of rehabilitation services remotely since such patients lack any triggering signal coming from their muscle tissue. As a result, this device will reduce the burden on the NHS and the time invested by doctors. Additionally, it would help keep patients away from the contaminated hospital environment. Since it reduces the cost of care considerably, the solution will be sustainable beyond the pandemic, saving journey times and resulting in a good uptake of video-based teleconsultation. Such a tele-rehab technique will satisfy the hospital’s commitment to reducing its carbon footprint. The PhD Student will research and present a clinical case where a patient has limited DOF and joint strength in a given joint. Besides students also research different types of rehabilitation exercises for neuromuscular disorders, need to explore different types of soft actuators for human joint movements, and understand and identify their limitations and advantages. They will work on soft actuator modelling and manufacturing the prototype and signal processing on biosensors such as EEG and EMG for human activity recognition and control. The project has an exciting opportunity to carry out collaborative research work with the University of Essex and East Kent Hospitals University NHS Foundation Trust.