Feasibility of interleaved stimulation in peroneal stimulators: the effect of surface electrode configuration on foot kinematics

Abstract

A peroneal stimulator is a neuroprosthetic device based on functional electrical stimulation (FES). It induces ankle dorsiflexion in the swing phase in patients affected by foot drop [1]. A well-known limitation of FES is the increased muscle fatigability due to the non-physiological motor unit (MU) activation. A possible approach to limit this issue is to interleave electrical pulses delivered to the muscle ( mstim ) and to the nerve ( nstim ) to distribute the stimulation among different MU groups [2], thus reducing the activation frequency of each group (interleaved stimulation). The practical applicability of this approach to peroneal stimulators requires that both mstim and nstim induce foot dorsiflexions (DFs) with minimal in-/eversion and ab-/adduction for all the knee angles within the swing phase. While for mstim knee rotations are not critical, changes in knee angle may modify the relative position between nstim electrodes and peroneal nerve, resulting in foot rotations outside the sagittal plane. Here we investigate the effect of nstim electrode positioning on induced foot rotations to identify the configuration leading to foot DF maximally similar to the voluntary ones for the entire knee ROM. We compared the electrically-induced foot movement for five nstim electrode positioning and four knee joint angles (0°, 20°, 40° and 60°, being 0° full knee extension) in ten healthy subjects. A grid of five cathodes was placed between fibular head and the tibial tuberosity (Fig. 1A) while a large anode (56 cm 2 ) was located on the patella. For each knee angle, electrical stimulation was delivered at the five stimulation sites (pulse duration: 200 µs, frequency: 20 Hz) and foot rotations were quantified with an IMU positioned under the foot. Foot in-/eversion and ab-/adduction were compared with those measured during voluntary DFs at each angle Fig. 1 B shows the ab-/adduction (Y-axes) and in-/eversion (X-axes) angles for each stimulation site (color-coded) and for the four knee angles in a representative subject. Red-dotted line delimits the acceptability area confining the angles obtained during voluntary DFs from all participants. All the points (stimulations) inside this area were regarded as producing a foot kinematic comparable to a voluntary DF. Figure c shows the number of subjects inside the acceptability area for all the stimulation configurations and knee angles. Collectively, configurations A and C included 7 out of 9 subjects, whilst for configurations D and E no subjects were inside the acceptability area for all knee angles. We identified a common nstim electrode configuration that induced, in the majority of the participants, foot (DF) with minimal eversion and inversion for the knee angles within the swing phase. This is a necessary condition to ensure the applicability of stimulation strategies based on distributed nerve/muscle stimulation (interleaved stimulation) aimed at reducing muscle fatigue in FES-based foot drop treatments.