ELECTRICAL NEUROMODULATION DURING ROBOT-ASSISTED STEPPING IN HUMANS WITH SPINAL CORD INJURY Matthias J. Krenn 1,2 1 Department of Neurobiology and Anatomical Sciences, University of Mississippi Medical Center, Jackson, MS, USA, 2 Center for Neuroscience and Neurological Recovery, Methodist Rehabilitation Center, Jackson, MS, USA mkrenn@umc.edu Abstract⎯ Electrical neuromodulation using trans- cutaneous spinal stimulation can modify the spinal motor output. In recent years, proof-of-principal stud- ies have shown the benefits of this intervention to re- cover locomotor functions. Here, we assess changes of joint torques during stimulation over a wide range of stimulation frequencies (1 – 100 Hz). The presented example shows high susceptibility to the external input by modifying stepping patterns during robot-assisted treadmill training. Keywords⎯ Transcutaneous spinal stimulation, neu- romodulation, sensorimotor integration, spinal cord in- jury, robot-assisted gait training Introduction Spinal cord injury (SCI) is a devastating neurological condition that affects the interactions between su- praspinal structures and the spinal cord below the le- sion. It results in partial or complete loss of volitional and postural control of movements associated with im- paired sensorimotor integration. The ensuing muscle weakness is often accompanied by spastic motor be- haviors, such as increased muscle tone (hypertonia), hyperactive reflexes (hyperreflexia), and clonus, as well as involuntary muscle contractions (spasms) and improper muscle coordination (dyssynergia) [1, 2]. New developments in electrical neuromodulation with transcutaneous (TSS) spinal stimulation show prom- ise for improving walking in people with SCI [3–5]. The underlying premise of TSS interventions is that the generated afferent input modifies the excitability of the lumbosacral network to either augment appropriate or suppress pathophysiologic spinal motor output [4, 6]. Here, we address the impact of TSS frequency from 1 up to 100 Hz on locomotor pattern in people with in- complete SCI. Methods Robot-assisted treadmill stepping The participant was first instrumented for EMG record- ing and TSS stimulation (see below). After determining the stimulation thresholds, the subject was placed in the bodyweight support harness and fitted into the ro- botic gait orthosis (Lokomat Pro V4, Hocoma AG, Volketswil, CH). The Lokomat (Fig. 1A) was used in a research mode, which provided real-time analog data output. This device controls leg movement towards a predefined trajectory of a physiological gait pattern by controlling the hip and knee joint torques of the exo- skeleton. A cascaded control system (Fig. 1B) inte- grates a first-order impedance controller (proportional- Figure 1. (A) Components of the robot-assisted gait exoskeleton, Lokomat Pro. (B) Cascaded control struc- ture. Primary angle controller based on desired (qdes) and measured actual (qact) angles; secondary torque control loop based generates the actuating torque (Tctr) of the exoskeleton; actual torque (Tact) measured at the actuators. The disturbance torques by a participant (Tpat) and treadmill (Ttm). Modified from Jezernik et al. (2003) [7]. Proc. Annual Meeting of the Austrian Society for Biomedical Engineering 2021 DOI: 10.3217/978-3-85125-826-4-10 CC BY Published by Verlag der TU Graz Graz University of Technology