Cable-Driven Robots in Physical Rehabilitation: From Theory to Practice

Introduction

Stroke is the leading cause of disability in the world and leaves a significant number of individuals with motor and cognitive deficits (Tucan et al., 2019; Gonçalves & Krebs, 2017). It can degrade upper and lower limbs functions and rehabilitation training is the most effective way to reduce these motor impairments in stroke patients. Robots can be suitable for this purpose since they can train patients for long durations with precision (Hatem et al., 2016). The robotic therapy assists the patient during training and can offer several advantages over conventional therapy as pointed out in Pennycott et al. (2012).

After a stroke the recovery of the movements is related to neural plasticity, which involves developing new neuronal interconnections, acquiring new functions and compensating for impairment. To promote neural plasticity and consequently movement’s recovery, the use of robots on rehabilitation can provide an intensive, repetitive and task-oriented training which has proven effective for the movements learning (Duret et al., 2019; Takahashi et al., 2016; Takeuchi & Izumi, 2013; Colombo et al., 2012; Kuznetsov et al., 2013; Maciejasz et al., 2014, Dzahir & Yamamoto, 2014).

Repeated movements of human limbs can help the patient to regain function of the injured limb. This process involves repetitive movements that must be performed several times at various speeds. The robots are, generally, more effective and efficient in performing these exercises than humans offering longer and more frequent therapy sessions for a patient and, consequently, they are attracting significant research interest. Additionally, a self-treatment with remote supervision of several patients by a single therapist professional can be implemented using these devices (Stein, 2012; Reinkensmeyer and Boninger, 2011).

Moreover, these systems can be coupled with virtual reality simulators, enhancing conventional physical therapy by recording patient information like position, force, time and velocity. The rehabilitation process involves other activities like diagnosis and prescription of the treatment to stimulate the recovery processes.

The recorded data by the robot can be achieved and compared through patient sessions to follow the progress (or even between different patients to aid in diagnosis), providing an objective measure of the progress and outcomes of therapy (Matin & Cazorla, 2019; Ceccarelli et al., 2010, Gonçalves et al., 2015).

These measures can offer a sensitive, accurate and time-efficient approach for the assessment of sensorimotor function after neurological impairment compared to standard clinical assessments. Besides that, motor functions of stroke patients are frequently related and evaluated using drawing/tracing tasks which are simpler with robotic systems (Takebayashi et al., 2018; Dipietro et al., 2007; Hussain et al., 2017). Additionally, the use of robotic solutions has strong potential for reducing treatment costs, which are becoming less affordable due to the aging of the population (Gassert & Dietz, 2018; Stein, 2012).

Different types of robotic architectures have been developed and applied in the rehabilitation of both upper and lower human limbs like industrial robots, exoskeletons, and cable-driven robots. In general, robotic structures used in rehabilitation are industrial robots or a new structure specifically designed or adapted to reproduce rehabilitation movements (Gonçalves et al., 2015; Gonçalves et al., 2019).

Industrial robots, usually have rigid structures, are fast, but several safety issues must be considered when applying them to rehabilitation (Mao et al., 2015). Exoskeletons for rehabilitation deal with a huge variability of human limbs shape and dimensions. In this way, matching the human and robot joints rotation centers is a very difficult task (Beyl et al., 2009).