A Robot-Driven Computational Model for Estimating Passive Ankle Torque With Subject-Specific Adaptation

An Online Estimating Framework for Ankle Actively Exerted Torque under Multi-DOF Coupled Dynamic Motions via sEMG

Ankle rehabilitation robots can offer tailored rehabilitation training, and facilitate the functional recovery of patients. Accurate estimation of the actively exerted torque from the ankle joint complex (AJC) can increase the engagement of patients during rehabilitation training. Given the three degrees of freedom (DOFs) of AJC and its coupled motion, it becomes essential to accurately estimate the actively exerted torque under multi-DOF. This work introduces an estimation framework that includes the Hill-based sEMG-force model, the ankle musculoskeletal dynamic decoupling model, and the parameter identification-calibration strategy. The Hill-based sEMG-force model estimates the force generated by individual muscles involved in AJC; The parameter identification-calibration strategy combined with pre-experiment identifies unknown variables in the ankle musculoskeletal dynamic decoupling model; Finally, the musculoskeletal dynamic decoupling model relates the muscle forces to the AJC's actively exerted torque. The musculoskeletal dynamic decoupling model combines anatomical and biomechanical features, enabling parameters derived from a single DOF pre-experiment through identification-calibration strategy to be applicable in multi-DOF dynamic motion. To evaluate the estimation performance of the framework, experiments were conducted in various directions involving both single and multiple DOFs. The results show that the proposed framework can estimate the actively exerted torque with a normalized root mean square error (NRMSE) of 10.29% ± 2.86% (mean ± SD) for torque estimation under a single DOF, and NRMSE of 11.35% ± 4.51% under multiple DOFs, compared to the actual measured values. This framework can improve human-robot interaction training and improve the effectiveness of robot-assisted ankle rehabilitation training. It can also provide accurate neuro-information and joint torque data for medical teams, which can lead to early diagnosis of diseases and patient-specific treatment protocols.

Augmentation-Consistent Clustering Network for Diabetic Retinopathy Grading with Fewer Annotations

Diabetic retinopathy (DR) is currently one of the severe complications leading to blindness, and computer-aided, diagnosis technology-assisted DR grading has become a popular research trend especially for the development of deep learning methods. However, most deep learning-based DR grading models require a large number of annotations to provide data guidance, and it is laborious for experts to find subtle lesion areas from fundus images, making accurate annotation more expensive than other vision tasks. In contrast, large-scale unlabeled data are easily accessible, becoming a potential solution to reduce the annotating workload in DR grading. Thus, this paper explores the internal correlations from unknown fundus images assisted by limited labeled fundus images to solve the semisupervised DR grading problem and proposes an augmentation-consistent clustering network (ACCN) to address the above-mentioned challenges. Specifically, the augmentation provides an efficient cue for the similarity information of unlabeled fundus images, assisting the supervision from the labeled data. By mining the consistent correlations from augmentation and raw images, the ACCN can discover subtle lesion features by clustering with fewer annotations. Experiments on Messidor and APTOS 2019 datasets show that the ACCN surpasses many state-of-the-art methods in a semisupervised manner.

A New Trajectory Determination Method for Robot-Assisted Ankle Ligament Rehabilitation

Keeping ligament strain at an appropriate range is beneficial for avoiding unexpected injuries and enhancing treatment efficacy. This study proposes a new trajectory determination method specifically for the robot-assisted ankle ligament rehabilitation. The input of this method is a set of strain constraints of certain ligaments and the output is the detailed training trajectory. Simulations were conducted with two cases (one-ligament injury and three-ligaments injury). While this method has not been experimentally tested, on condition of an accurate ligament kinematics assessment, ligament strain can be guaranteed to be within the specified range following the derived trajectory. This method can help design injury-specific treatment protocols and has potential in improving the effectiveness of robot-assisted ankle rehabilitation. Future work will verify the validity and the practicality, and consider the improvement of the method.

Research on an Ankle Joint Auxiliary Rehabilitation Robot with a Rigid-Flexible Hybrid Drive Based on a 2-S'PS' Mechanism

An ankle joint auxiliary rehabilitation robot has been developed, which consists of an upper platform, a lower platform, a dorsiflexion/plantar flexion drive system, a varus/valgus drive system, and some connecting parts. The upper platform connects to the lower platform through a ball pin pair and two driving branch chains based on the S′PS′ mechanism. Although the robot has two degrees of freedom (DOF), the upper platform can realize three kinds of motion. To achieve ankle joint auxiliary rehabilitation, the ankle joint of patients on the upper platform makes a bionic motion. The robot uses a centre ball pin pair as the main support to simulate the motion of the ankle joint; the upper platform and the centre ball pin pair construct a mirror image of a patient's foot and ankle joint, which satisfies the human body physiological characteristics; the driving systems adopt a rigid-flexible hybrid structure; and the dorsiflexion/plantar flexion motion and the varus/valgus motion are decoupled. These structural features can avoid secondary damage to the patient. The rehabilitation process is considered, and energy consumption of the robot is studied. An experimental prototype demonstrates that the robot can simulate the motion of the human foot.

Hierarchical Compliance Control of a Soft Ankle Rehabilitation Robot Actuated by Pneumatic Muscles

Traditional compliance control of a rehabilitation robot is implemented in task space by using impedance or admittance control algorithms. The soft robot actuated by pneumatic muscle actuators (PMAs) is becoming prominent for patients as it enables the compliance being adjusted in each active link, which, however, has not been reported in the literature. This paper proposes a new compliance control method of a soft ankle rehabilitation robot that is driven by four PMAs configured in parallel to enable three degrees of freedom movement of the ankle joint. A new hierarchical compliance control structure, including a low-level compliance adjustment controller in joint space and a high-level admittance controller in task space, is designed. An adaptive compliance control paradigm is further developed by taking into account patient’s active contribution and movement ability during a previous period of time, in order to provide robot assistance only when it is necessarily required. Experiments on healthy and impaired human subjects were conducted to verify the adaptive hierarchical compliance control scheme. The results show that the robot hierarchical compliance can be online adjusted according to the participant’s assessment. The robot reduces its assistance output when participants contribute more and vice versa, thus providing a potentially feasible solution to the patient-in-loop cooperative training strategy.