Brain-machine interfaces (BMIs) provide a new assistive strategy aimed at restoring mobility in severely paralyzed patients. Yet, no study in animals or in human subjects has indicated that long-term BMI training could induce any type of clinical recovery. Eight chronic (3–13 years) spinal cord injury (SCI) paraplegics were subjected to long-term training (12 months) with a multi-stage BMI-based gait neurorehabilitation paradigm aimed at restoring locomotion. This paradigm combined intense immersive virtual reality training, enriched visual-tactile feedback, and walking with two EEG-controlled robotic actuators, including a custom-designed lower limb exoskeleton capable of delivering tactile feedback to subjects. Following 12 months of training with this paradigm, all eight patients experienced neurological improvements in somatic sensation (pain localization, fine/crude touch, and proprioceptive sensing) in multiple dermatomes. Patients also regained voluntary motor control in key muscles below the SCI level, as measured by EMGs, resulting in marked improvement in their walking index. As a result, 50% of these patients were upgraded to an incomplete paraplegia classification. Neurological recovery was paralleled by the reemergence of lower limb motor imagery at cortical level. We hypothesize that this unprecedented neurological recovery results from both cortical and spinal cord plasticity triggered by long-term BMI usage.
Spinal Cord Injury (SCI) rehabilitation remains a major clinical challenge, especially in cases involving chronic complete injury. Clinical studies using body weight support systems1,2, robotic assistance1,2,3,4, and functional electrostimulation of the leg5,6 have proposed potential solutions for assisting SCI patients in walking7,8. Yet, none of these approaches have generated any consistent clinical improvement in neurological functions, namely somatosensory (tactile, proprioceptive, pain, and temperature) perception and voluntary motor control, below the level of the spinal cord lesion.
Since the first experimental demonstrations in rats9, monkeys10,11, and the subsequent clinical reports in humans12,13,14, brain-machine interfaces (BMIs) have emerged as potential options to restore mobility in patients who are severely paralyzed as a result of spinal cord injuries (SCIs) or neurodegenerative disorders15. However, to our knowledge, no study has suggested that long-term training associating BMI-based paradigms and physical training could trigger neurological recovery, particularly in patients clinically diagnosed as having a complete SCI. Yet, in 60–80% of these “complete” SCI patients, neurophysiological assessments16,17 and post-mortem anatomical18 studies have indicated the existence of a number of viable axons crossing the level of the SCI. This led some authors to refer to these patients as having a “discomplete” SCI17 and predict that these remaining axons could mediate some degree of neurological recovery.
For the past few years, our multidisciplinary team has been engaged in a project to implement a multi-stage neurorehabilitation protocol – the Walk Again Neurorehabilitation (WA-NR) – in chronic SCI patients. This protocol included the intensive employment of immersive virtual-reality environments, combining training on non-invasive brain-control of virtual avatar bodies with rich visual and tactile feedback, and the use of closed-loop BMI platforms in conjunction with lower limb robotic actuators, such as a commercially available robotic walker (Lokomat, Hocoma AG, Volketswil, Switzerland), and a brain-controlled robotic exoskeleton, custom-designed specifically for the execution of this project.
Originally, our central goal was to explore how much such a long-term BMI-based protocol could help SCI patients regain their ability to walk autonomously using our brain-controlled exoskeleton. Among other innovations, this device provides tactile feedback to subjects through the combination of multiple force-sensors, applied to key locations of the exoskeleton, such as the plantar surface of the feet, and a multi-channel haptic display, applied to the patient’s forearm skin surface.
Unexpectedly, at the end of the first 12 months of training with the WA-NR protocol, a comprehensive neurological examination revealed that all of our eight patients had experienced a significant clinical improvement in their ability to perceive somatic sensations and exert voluntary motor control in dermatomes located below the original SCI. EEG analysis revealed clear signs of cortical functional plasticity, at the level of the primary somatosensory and motor cortical areas, during the same period. These findings suggest, for the first time, that long-term exposure to BMI-based protocols enriched with tactile feedback and combined with robotic gait training may induce cortical and subcortical plasticity capable of triggering partial neurological recovery even in patients originally diagnosed with a chronic complete spinal cord injury.
Eight paraplegic patients, suffering from chronic (>1 year) spinal cord injury (SCI, seven complete and one incomplete, see Fig. 1A, Supplementary Methods Inclusion/exclusion Criteria), were followed by a multidisciplinary rehabilitation team, comprised of clinical staff, engineers, neuroscientists, and roboticists, during the 12 months of 2014. Our clinical protocol, which we named the Walk Again Neurorehabilitation (WA-NR), was approved by both a local ethics committee (Associação de Assistência à Criança Deficiente, Sao Paulo, Sao Paulo, Brazil #364.027) and the Brazilian federal government ethics committee (CONEP, CAAE: 13165913.1.0000.0085). All research activities were carried out in accordance with the guidelines and regulations of the Associação de Assistência à Criança Deficiente and CONEP. Each participant signed written informed consent before enrolling in the study. The central goal of this study was to investigate the clinical impact of the WA-NR, which consisted of the integration between traditional physical rehabilitation and the use of multiple brain-machine interface paradigms (BMI). This protocol included six components: (1) an immersive virtual reality environment in which a seated patient employed his/her brain activity, recorded via a 16-channel EEG, to control the movements of a human body avatar, while receiving visuo-tactile feedback; (2) identical interaction with the same virtual environment and BMI protocol while patients were upright, supported by a stand-in-table device; (3) training on a robotic body weight support (BWS) gait system on a treadmill (Lokomat, Hocoma AG, Switzerland); (4) training with a BWS gait system fixed on an overground track (ZeroG, Aretech LLC., Ashburn, VA); (5) training with a brain-controlled robotic BWS gait system on a treadmill; and (6) gait training with a brain-controlled, sensorized 12 degrees of freedom robotic exoskeleton (seeSupplementary Material).