Toward a whole-body neuroprosthetic

Abstract

Brain–machine interfaces (BMIs) hold promise for the restoration of body mobility in patients suffering from devastating motor deficits caused by brain injury, neurological diseases, and limb loss. Considerable progress has been achieved in BMIs that enact arm movements, and initial work has been done on BMIs for lower limb and trunk control. These developments put Duke University Center for Neuroengineering in the position to develop the first BMI for whole-body control. This whole-body BMI will incorporate very large-scale brain recordings, advanced decoding algorithms, artificial sensory feedback based on electrical stimulation of somatosensory areas, virtual environment representations, and a whole-body exoskeleton. This system will be first tested in nonhuman primates and then transferred to clinical trials in humans.

Introduction

Brain–machine interfaces (BMIs) (Andersen et al., 2004, Birbaumer and Cohen, 2007, Fetz, 2007, Lebedev and Nicolelis, 2006, Nicolelis, 2001, Nicolelis and Lebedev, 2009, Schwartz et al., 2006, Wessberg et al., 2000) offer a translational solution to the problem of restoring mobility to millions of people who suffer from paralysis caused by neurological injuries, neurodegenerative diseases, or limb loss (Paddock, 2009). Only limited treatment options are available to these patients, and often their condition cannot be improved or ameliorated (Dobkin and Havton, 2004, Fouad and Pearson, 2004). BMIs hold promise to revolutionize the treatment of paralysis, as they strive to repair the damaged neural circuitry by bypassing the site of the lesion and establishing direct neural control of artificial tools by the activity of intact brain areas, such as the primary motor cortex (M1), which in many cases remain capable of generating motor commands despite being disconnected from the body effectors (Mattia et al., 2009). BMI research has expanded rapidly during the past decade (Lebedev and Nicolelis, 2006, Nicolelis and Lebedev, 2009), generating high expectations for potential clinical applications. Proof-of-concept BMIs have been tested in rodents (Chapin et al., 1999), nonhuman primates (Carmena et al., 2003, Moritz et al., 2008, Taylor et al., 2002, Velliste et al., 2008, Wessberg et al., 2000), and human subjects (Allison et al., 2007, Birbaumer and Cohen, 2007, Hochberg et al., 2006, Kennedy and Bakay, 1998, Patil et al., 2004, Pfurtscheller and Neuper, 2006). BMI systems developed at the Duke University Center for Neuroengineering (DUCN) during the past 12 years have made it possible to control many motor functions by neuronal ensemble activity recorded with chronic implants, ranging from arm reaching and grasping movements (Carmena et al., 2003, Lebedev et al., 2005, Wessberg et al., 2000) to bipedal locomotion (Cheng, Hyon, et al., 2007, Fitzsimmons et al., 2009). Moreover, recently we have demonstrated for the first time brain–machine–brain interfaces (BMBIs) that incorporate somatosensory feedback loops that transmit information from the actuator to the brain (O'Doherty et al., 2009, O'Doherty et al., 2010). These developments have put us in the position to develop the first whole-body neuroprosthetic for severely paralyzed patients.