[ARTICLE] An Assist-as-Needed Controller for Passive, Assistant, Active, and Resistive Robot-Aided Rehabilitation Training of the Upper Extremity – Full Text

Abstract

Clinical studies have demonstrated that robot-involved therapy can effectively improve the rehabilitation training effect of motor ability and daily behavior ability of subjects with an upper limb motor dysfunction. This paper presents an impedance-based assist-as-needed controller that can be used in robot-aided rehabilitation training for subjects with an upper extremity dysfunction. Then, the controller is implemented on an end-effector upper extremity rehabilitation robot which could assist subjects in performing training with a spatial trajectory. The proposed controller enables subjects’ arms to have motion freedom by building a fault-tolerant region around the rehabilitation trajectory. Subjects could move their upper limb without any assistance within the fault-tolerant region while the robot would provide assistance according to the subjects’ functional ability when deviating from the fault-tolerant region. Besides, we also put forward the stiffness field around the fault-tolerant region to increase the robot’s assistance when subjects’ hand is moving outside the fault-tolerant region. A series of columnar rigid walls would be constructed in the controller according to the subjects’ functional ability, and the stiffness of the wall increases as the motion performance deteriorates. Furthermore, the controller contains five adjustable parameters. The controller would show different performances by adjusting these parameters and satisfy the requirement of robot-aided rehabilitation training at different rehabilitation stages such as passive, assistant, active, and resistant training. Finally, the controller was tested with an elderly female participant with different controller parameters, and experimental results verified the correctness of the controller and its potential ability to satisfy the training requirements at different rehabilitation stages. In the close future, the proposed controller in this work is planned to be applied on more subjects and also patients who have upper limb motor dysfunctions to demonstrate performance of the controller with different parameters.

1. Introduction

In recent years, the number of subjects with upper extremity motor dysfunctions caused by stroke, spinal cord injuries, and accidents has been dramatically increasing year by year [1], severely limiting their motor and activities of daily living (ADL) abilities. This not only brings physical inconvenience to the subjects but also brings financial and mental burden to their families. Research results have showed that compared with the traditional manual rehabilitation therapy, robot-aided rehabilitation training could carry out high-intensity repetitive and task-oriented training tasks [2], and clinical research studies have also shown that robot-aided rehabilitation therapy could effectively improve the rehabilitation effect on patients’ upper extremity motor abilities [3,4]. To satisfy the patient’s requirements at different rehabilitation stages, different types of upper limb rehabilitation robots have been developed, which can be divided into two categories [5]: the end-effector type [6,7] and the exoskeleton type [8,9]. Generally, exoskeleton type upper limb rehabilitation robots could implement rehabilitation training of a single joint; however, additional interactive forces and torques between the patient’s arm and the robot would occur [10]. Compared with the traditional rehabilitation training methods, the end-effector type upper limb rehabilitation robot has better superiority during upper limb rehabilitation training [11]. Clinical comparative trial results have also showed that compared with the exoskeleton upper limb rehabilitation robot, intervention with the end-effector rehabilitation robot was more effective in terms of the active participation of patients with mild and severe stroke [12].According to the Brunstorm theory, patients with an upper limb motor disorder in the early recovery stage experience difficulties in moving their arm, and therapists or robots are required to carry out repeated passive training. Therefore, a stiff position controller is required to be applied in such passive training. The rehabilitation effect of the passive training method would be not great with the recovery of the patient’s motor ability because the active participation of the patients will be ignored in those position controllers. For patients with partial motor control ability, robot intervention to a minimum extent could stimulate and promote the brain neuroplasticity of patients [13,14]. Studies have shown that task-oriented rehabilitation training is only effective when it is associated with task-oriented movements involving effort by the subject actively [15]. Therefore, a controller needs to be developed to provide necessary assistance or correction to complete rehabilitation training tasks according to the patient’s functional ability. Controllers with such characteristics are called assist-as-needed (AAN) controllers, also known as assistance or corrective controllers. At present, AAN controllers have been used in lower limb rehabilitation [16], finger rehabilitation [17], and upper limb rehabilitation [18] to stimulate the patient’s active participation during rehabilitation training. The results of a clinical trial proved that an upper limb rehabilitation robot with an AAN controller could provide intensive and repeated rehabilitation training and could also promote the active participation and motion performance of patients, promoting the motor recovery [19].Regarding robot-aided training with the AAN characteristic, impedance control is often applied to assist patients during robot-aided rehabilitation training tasks [20]—the schematic diagram is shown in Figure 1. Clinical studies on stroke patients [21,22] showed that rehabilitation robots with the impedance control strategy have a better effect on reducing the upper limb muscle strength and improving the motor function of patients than robot-assisted therapy that simply mimics traditional therapy [23,24,25]. Although the impedance control strategy allows a patient’s arm to deviate from the training trajectory, the patient’s spatial freedom is limited. Assistance from the robot is always given as a restoring force to assist the patients once they are deviating from the planned trajectory. Therefore, scholars have developed an impedance-based AAN strategy to increase the patient’s freedom, which can be mainly divided into two categories: the band-type controller and the window-type controller. The window-type controller defines a predetermined trajectory and a moving window that allows the patient’s arm to move freely. When the patient’s arm lies outside the moving window, the patient’s arm would be corrected into the moving window by the robot’s assistance force [26,27]. Another band-type controller is based on the fault-tolerant region (FTR), which constructs a fault-tolerant band that allows the patient to move freely by establishing the inner and outer boundary of the FTR. The patient’s spatial freedom is limited in the window-type AAN controller, while the band-type AAN controller allows the patient to have complete freedom and choose his/her own trajectory without being assisted by the robot. The patient’s arm has complete freedom when moving in the FTR, while the robot will apply an assistance force to assist the patient’s arm to finish the rehabilitation training task when deviating from FTR. Research studies have also been conducted to explore increasing the patient’s freedom. Ying et al. [28] first developed a rope-driven exoskeleton rehabilitation robot and then added a fault-tolerant area around the predetermined trajectory to increase the patient’s spatial freedom. Hamed et al. [29] validated the feasibility by adding a plane channel around the predetermined trajectory based on a two-link rehabilitation robot. Although the band-type AAN controller has shown positive results in clinical trials, its tendency or ‘‘slacking’’ for the patient to rely heavily on the assistance force has been recorded [30]. The robot will not apply any assistance or correction as long as the patient’s arm is moving within the FTR, which will greatly reduce the active participation of patients. A velocity curve could be added in the FTR which will build a virtual movement in the FTR. When the patient’s arm is stagnant or moving too slowly in the FTR, the controller will generate a moving wall to push the patient’s arm to catch up with the virtual movement to complete the training task and avoid the slacking.

Figure 1. The diagram scheme of the impedance control; the black curve is the desired trajectory, while the blue curve is the actual trajectory.

To better adjust the assistance force of the AAN controller, it is necessary to evaluate the patient’s real-time motor ability or functional ability. One can conclude that there are mainly two classifications for the evaluation of a patient’s functional ability based on the current literatures: the sensor-based [31,32] and model-based [33] evaluation techniques. Although there is more of a theoretical basis for the evaluation of upper limb motor ability for model-based methods, the modeling results are difficult to use to accurately evaluate the patient’s true motion intention, while sensor-based evaluation methods can more intuitively express the patient’s motion state. Some studies have already used the position deviation between the actual trajectory of a patient’s hands and the expected trajectory which was calculated to represent the functional ability of patients during rehabilitation training [34].[…]

Continue