Restoring voluntary actions to paralysed patients is an ambition of neural-interface research. A study shows that people with tetraplegia can use brain control of a robotic arm to reach and grasp objects. See Letter p.372
Most of us take for granted our effortless ability to interact with objects. When we are thirsty and reach for a cup, electrical signals stream from the brain through the spinal cord, instructing our muscles to move. However, a disruption of the nerve pathways along which these signals travel can cause paralysis, with devastating consequences for the person's quality of life. So there is growing interest in technologies that allow the brain to bypass nerve injuries and communicate directly with the environment. Neural-interface systems, also known as brain–machine interfaces, detect electrical signals in the brain and use them to control external assistive devices. The first results from a clinical trial of 'BrainGate', a neural interface that enabled a patient paralysed by a spinal-cord injury to move a computer cursor, were published1 in 2006. On page 372 of this issue, Hochberg et al.2 now report that two people with long-standing paralysis can control the reaching and grasping actions of a robotic arm using BrainGate. One of the participants was even able to drink from a bottle using the robotic arm, something she had not been able to do with her own limb since a stroke nearly 15 years ago.
To access brain signals, BrainGate uses thin silicon electrodes surgically inserted a few millimetres into the primary motor cortex, a part of the brain that controls movements. Remarkably, neurons in this area responded when the patients imagined controlling the robotic arm, although both of them had lost the use of their limbs many years earlier. During a calibration phase, the researchers constructed a 'decoder' that translated participants' intentions into three-dimensional movements and into a closing of the robotic hand. They then tested the participants' ability to reach for and grasp foam balls presented in front of them.
Although the speed and accuracy of the robot's movements fell well short of those of natural arm control, the participants successfully touched the foam balls on 49% to 95% of attempts across multiple sessions with two different robot designs. What's more, about two-thirds of successful reaches resulted in correct grasping. The authors further established the efficacy of brain control by one participant in a bottle-grasping and drinking task, demonstrating that a neural-interface system can perform actions that are useful in daily life.
Apart from being one of only a handful of studies to use indwelling electrodes for neural interfacing in humans, Hochberg and colleagues' work is notable in that one patient had had the implanted electrodes for more than five years. Although several techniques (such as electroencephalography) can record signals from the brain in a non-invasive manner, it is generally thought that electrodes positioned inside the brain convey more information. However, as well as the risks associated with surgery, a disadvantage of such implants is the potential for scar tissue to form around the electrodes, which can result in a deterioration of signal quality over time. The authors acknowledge that some deterioration had indeed occurred, but it is encouraging that useful signals could still be obtained five years later. Nevertheless, as many spinal-cord injuries occur at a young age and patients may live with their disabilities for many decades, further efforts to understand and control the tissue response to indwelling electrodes will be crucial for widespread clinical application of neural-interface systems.
The current study also underlines the importance of basic research in driving translational advances. At a time when experimentation using non-human primates is increasingly controversial, it is worth noting that the results reported by Hochberg et al. draw directly on previous neural-interface demonstrations in monkeys3,4,5,6 and on decades of basic research into the control of arm movements (Fig. 1). The upper limb of primates is a uniquely versatile tool, and its evolution involved profound changes to the motor structures of the brain and their descending connections that are not shared with other mammals such as mice and rats7. Further understanding of how distributed populations of neurons in the brain and the spinal cord cooperate during dexterous manipulation of objects will doubtless inform the development of improved neural interfaces for artificial limbs.
Paralysis resulting from nervous-system injury is a multifaceted condition, with many aspects that affect quality of life. Nevertheless, patients consistently rate regaining arm and hand function as a top priority8. So, although robotic arms may be of practical assistance, restoring movements of the patients' own limbs should remain the ultimate goal. Future neural-interface systems may help to achieve this, if they can be coupled to functional electrical stimulation of muscles9,10 or the spinal cord11. In addition, damage to sensory pathways may require artificial sensation to be relayed to the brain before fully naturalistic movements can be restored12. Furthermore, all this should preferably be performed by wireless implants that do not physically breach the skin. Encouraging progress is being made on all these fronts.
Ultimately, the greatest obstacle to clinical applications of neural interfaces may come not from science or engineering, but from economics. The original BrainGate clinical trial was initiated by a US company (Cyberkinetics Neurotechnology Systems) that ceased operations in 2009. Fortunately, a new clinical trial is in progress, administered by the Massachusetts General Hospital in Boston. It remains to be seen whether a neural-interface system that will be of practical use to patients with diverse clinical needs can become a commercially viable proposition. Nevertheless, the delight of a participant in Hochberg and colleagues' study as she succeeds in drinking from a bottle for the first time in years (see Supplementary Movie 4 that accompanies the paper2) should act as a powerful incentive for all in the field to address these challenges.
Hochberg, L. R. et al. Nature 442, 164–171 (2006).
Hochberg, L. R. et al. Nature 485, 372–375 (2012).
Wessberg, J. et al. Nature 408, 361–365 (2000).
Serruya, M. D., Hatsopoulos, N. G., Paninski, L., Fellows, M. R. & Donoghue, J. P. Nature 416, 141–142 (2002).
Taylor, D. M., Tillery, S. I. & Schwartz, A. B. Science 296, 1829–1832 (2002).
Velliste, M., Perel, S., Spalding, M. C., Whitford, A. S. & Schwartz, A. B. Nature 453, 1098–1101 (2008).
Lemon, R. N. Annu. Rev. Neurosci. 31, 195–218 (2008).
Anderson, K. D. J. Neurotrauma 21, 1371–1383 (2004).
Moritz, C. T., Perlmutter, S. I. & Fetz, E. E. Nature 456, 639–642 (2008).
Ethier, C., Oby, E. R., Bauman, M. J. & Miller, L. E. Nature 485, 368–371 (2012).
Zimmermann, J. B., Seki, K. & Jackson, A. J. Neural Eng. 8, 054001 (2011).
O'Doherty, J. E. et al. Nature 479, 228–231 (2011).
Evarts, E. V. J. Neurophysiol. 27, 152–171 (1964).
Humphrey, D. R., Schmidt, E. M. & Thompson W. D. Science 170, 758–762 (1970).
Georgopoulos, A. P., Kalaska, J. F., Caminiti, R. & Massey, J. T. J. Neurosci. 2, 1527–1537 (1982).
Nordhausen, C. T., Rousche, P. J. & Normann, R. A. Brain Res. 637, 27–36 (1994).
About this article
Modeling and Analysis of a Modular Multilegged Robot with Improved Fault Tolerance and Environmental Adaptability
Mathematical Problems in Engineering (2019)
Line Spectrum Representation of Neural and Myoelectric Signals for Brain-Machine Interface Applications
Proceedings of the ISCIE International Symposium on Stochastic Systems Theory and its Applications (2018)
Reduction of the foreign body response and neuroprotection by apyrase and minocycline in chronic cannula implantation in the rat brain
Clinical and Experimental Pharmacology and Physiology (2017)
Suppression of excitotoxicity and foreign body response by memantine in chronic cannula implantation into the rat brain
Brain Research Bulletin (2015)