Goal-oriented, subject adaptive, robot-assisted locomotor learning (GOALL)

The objective of this line of research is to identify robotic controllers capable of training subjects to modify kinematics and kinetics in desired ways with the goal of extending this approach to neurorehabilitation.

In this project, we wish to modulate trailing limb angle utilizing a robotic controller which applies pulses of torque at key moments in gait. In order to develop such a controller, we asked how sagittal plane joint torque is affected by two important gait parameters: gait speed and stride length.

We conducted an experiment with healthy control subjects instructed to walk on a treadmill at various speeds, and asked to modulate stride length via visual feedback.

Sagittal plane joint torques were extracted via an inverse dynamics analysis of instrumented treadmill and motion capture data.

We used a torque pulse approximation analysis to determine optimal timing and amplitude of torque pulses that approximate the SL-specific difference in joint torque profiles measured at different values of GS.

Our group analysis has generated a set of 16 pulse torque assistance profiles that is being experimentally tested using a gait training exoskeleton, with 2 active d.o.f., actuating about the hip and knee joints.

Healthy control subjects have undergone single stride exposures of the 16 pulse torque assistance profiles in which the effects on gait parameters such as hip extension and propulsive impulse have been measured.

Data stored in repository

De-identified subject data and processing files have been deposited in the Zenodo repository (DOI: 10.5281/zenodo.2547699).

Publications on this topic

R. L. McGrath, M. Pires-fernandes, B. Knarr, J. S. Higginson, and F. Sergi, “Toward goal-oriented robotic gait training : the effect of gait speed and stride length on lower extremity joint torques,” IEEE/RAS-EMBS Int. Conf. Rehabil. Robot., 2017.

R. L. McGrath, M. L. Ziegler, M. Pires-Fernandes, B. A. Knarr, J. S. Higginson, and F. Sergi, “The effect of stride length on lower extremity joint kinetics at various gait speeds,” PLoS One, vol. 14, no. 2, p. e0200862, Feb. 2019.


Goal Oriented Gait Training using a split-belt treadmill

Many stroke survivors suffer from hemiparesis, a condition that results in impaired walking ability. Walking ability is commonly assessed by walking speed, which is dependent on propulsive force both in healthy and stroke populations. Propulsive force is determined by two factors: ankle moment and the posture of the trailing limb during push-off. Ankle moment can be quantified by measuring the anterior ground reaction forces (AGRF) during treadmill walking, and posture of the trailing limb can be measured as the angle between the trailing limb and the vertical axis shown below.

Recent work has used robotic assistance strategies to modulate propulsive force with some success. However, robotic strategies are limited by their high cost and the technical
difficulty of fitting and operating robotic devices with stroke survivors in a clinical setting.

To address these limitations, we have developed a new protocol for goal-oriented gait training that utilizes a split belt treadmill to train both components of propulsive force generation, achieved by accelerating the treadmill belt of the trailing limb during push off. Belt accelerations require subjects to produce greater propulsive force to maintain their position on the treadmill and increases trailing limb angle through increased velocity of the accelerated limb.

Treadmill belt accelerations during intervention as a function of gait cycle

We hypothesized that accelerations would cause locomotor adaptation that would result in measurable after effects in the form of increased propulsive force generation. We tested our protocol (shown below) on healthy subjects at two levels of belt accelerations; a perceptible  7 m/s^2 acceleration group (N=9) and an imperceptible 2 m/s^2 group (N=10).

Experimental Protocol Schematic. Belt accelerations are shown on the y-axis, where epsilon signifies the magnitude of accelerations applied during intervention (either 2 m/s^2 or 7 m/s^2). A minute long “ramp” phase (green), in which accelerations linearly increased in magnitude from 0 to epsilon, was used to gradually introduce the intervention condition. The User-Driven speed conditions were conducted using a user-driven treadmill controller that modifies treadmill speed in response to changes in subjects 1) position on the treadmill (forward = increased speed, backward = decreased speed, 2) change in AGRF (increases = increased speed), and 3) change in step length (increases = increased speed). In the UDTC condition , maximum belt accelerations were capped at 0.2 m/s^2. Highlighted phases at the bottom signify periods in which kinematic marker data (used to assess TLA) were collected.

To evaluate the effects of our intervention on walking ability, we compared AGRF, Gait speed (GS) and TLA measured in the baseline 1 block, (BL1) to measurements collect in the baseline 2 block (BL2). To evaluate how change in walking behavior evolved over the course of BL2, we partitioned BL2 into 20 stride windows, and compared each window to BL1 (shown below).

Group level pre and post intervention change in AGRF, Gait speed (GS), and TLA. Each bar plot displays the mean and standard error in each bin. Astrics at the top of each plot display significant change between BL2 bins and BL1. Astrics at the bottom display significant change between denoted BL2 bins.

Overall, our paradigm significantly increased propulsive force generation in 78% of subjects in the Perceptible group (7 m/s^2) and 80% in the Imperceptible group (2 m/s^2), that translated to increases in gait speed in 66% and 50% of subjects respectively. Our results indicate the Perceptible intervention is a better candidate for continued research in training propulsion, as subjects in this group had more consistent after-effects, larger increases in all gait parameters tested, and sustained change in walking behavior following intervention.

Publications on this topic

A. J. Farrens, R. Marbaker, M. Liley, F. Sergi“Training propulsion during walking: adaptation to accelerations of the trailing limb”, In review, bioRXiv, March 2019.

Print Friendly, PDF & Email