REHABILITATION FIRST, TECHNOLOGY AS A TOOL

M. Petrarca¹, M. Bottoni¹, M. Favetta¹, G. Vasco¹, E. Castelli¹, S. Summa¹

¹"Bambino Gesù" Children's Hospital, Neurorehabilitation and Robotics, Rome, Italy

Background: Robotic wearable orthoses are permeating the rehabilitative processes to respond to the demands of re-education and support of standing and walking functions. Literature revisions on their taxonomy and usage partially address the rehabilitative issues or address only technical aspects of the human-machine interaction[1]. The literature considers mainly biomechanical coupling, wearability, intention detecting, functional coupling, energy autonomy, trajectory definition, and torque expectation. What is missing is the analysis of the scientific rules of interaction able to induce learning processes or support the function. Furthermore, over 30 years of instrumented gait analysis changed the interpretation of human locomotion organization, currently considered a complex dynamic function[2]. Testing technological solutions in pathological conditions let to consider the necessity to change the research paradigms in this field and the need for low-cost solutions[3].

Purpose: This study focuses on advanced motor control and learning hypotheses, reversing the current processes that focus on the technology first. This reverse has ethical and scientific reasons. The test of scientific perspectives is the aim of this study and necessitates an ad-hoc built knee robotic orthosis and controller. In this study, the prototype is verified on healthy participants to test the possibility of following and modifying natural walking.

Methods: We developed an ad-hoc orthosis with a controller that follows the equilibrium-point hypothesis[4]. Seven typically developed volunteers participated in the test of the orthosis controller (age: 22 – 65 years old, 5 females). Data were gathered through a 12-camera optoelectronic system (Vicon, MX, UK). Three experimental protocols tested the transparency and the gait pattern implicit modification using the orthosis during walking, stepping in place on a compliant platform with 6 DoF (DORIS, Mufy, IT)[4] and while walking on a treadmill. The orthosis was controlled in torque by current modulation and imposing phase-dependent target body references.

Results: The ad-hoc low-cost built bioinspired controller was efficient during unperturbed walking, perturbed walking, and stepping in place on DORIS. Knee flexion-extension time series gathered with gait analysis instruments overlap comparing them with steady gait during the use of the orthosis in transparency mode. Furthermore, while walking on the treadmill, it was possible to dynamically modify the gait kinematic, maintaining the general pattern configuration by changing the current driving the torque of the actuator and selectively changing the final target configuration only during the swing phase. The torque and final angle target worked subliminally concerning the participant's perception but were efficient in changing the maximum knee flexion angle.

Conclusions: From our preliminary data, the phase-dependent control on the final configuration through a force interaction realized the possibility of following the participants' behaviour transparently without any control on the trajectory of the movement. These results match the equilibrium point hypothesis[5].

Implications: Interacting with gait patterns through subliminal information opens the path for recovering, when possible, or supporting natural gait, avoiding dangerous interactions that can cause stumble and falls through a low-cost solution.

Funding acknowledgements: The EU Grant EUROBENCH supported this work.

Keywords:
Gait Recovery
Motor Control
Robotic Orthosis

Topics:
Research methodology, knowledge translation & implementation science
Innovative technology: robotics
Disability & rehabilitation

Did this work require ethics approval? Yes

Institution: Bambino Gesù Children's Hospital

Committee: Bambino Gesù Children's Hospital Ethical Committee

Ethics number: 1537/2018