康复学报 (Jun 2023)
Research Progress of Intelligent Evaluation and Virtual Reality Based Training in Upper Limb Rehabilitation afrer Stroke
Abstract
Stroke is the leading cause of disability worldwide, with approximately 85% of poststroke patients experiencing upper extremity dysfunction at first presentation and only 5% to 20% fully recovered at six months post-stroke. As a result, patients often experience difficulties in daily life, and improving the functional use of the upper extremities in poststroke patients is critical. Accurate assessment and training of motor function in stroke patients is essential. Integrating motor assessment with clinical assessment scales is vital to the comprehensive evaluation of upper limb function aspects fully. Markerless sensing techniques are used to assess typical motor abilities of the upper extremity in poststroke patients, such as range of motion, coordination, grip strength, or fine manual dexterity. Automated assessment of upper extremity motor function based on machine learning algorithms with markerless sensing techniques has focused on the Fugl-Meyer assessment of upper extremity (FMA-UE), Brunnstrom stages, and Wolf motor function test (WMFT) scales and has been proved with high-scoring accuracy and time efficiency. In addition, virtual reality (VR) training designed for neurorehabilitation can effectively promote recovery of upper limb function after stroke. The principles of neurorehabilitation associated with it include task-oriented practice, explicit feedback, increased difficulty, implicit feedback, variable practice, and mechanisms to promote the use of the affected limb. The effectiveness of VR training is the result of the advantage of the technology combined with neurorehabilitation principles to promote motor learning and functional recovery. VR rehabilitation systems can be divided into off-the-shelf commercial video game systems and custom virtual environment systems; touch and touchless interfaces; proximal movement (shoulder and elbow), distal movement (wrist and hand), and hybrid movement systems. Various VR rehabilitation systems for upper limb function in stroke have been developed at home and abroad, but there are shortcomings: 1) lack of patient-centered task-oriented personalized training plans; 2) less frequent use of the hand and arm as a whole in the research; 3) more subjective assessment of functional impairment and clinical outcome; 4) no close integration between rehabilitation assessment and treatment. Moreover, the intelligent assessment of upper limb motor function in stroke has focused on the FMA-UE, Brunnstrom, and WMFT scales, and the functional test for the hemiplegic upper extremity (FTHUE) scale, which is widely used in clinical practice, has not been studied. The FTHUE assessment scale and a set of rehabilitation programs based on the FTHUE scale are commonly used in clinical practice. Combining the FTHUE rehabilitation program with VR training can more effectively promote patients' motor function recovery. Future research can select the corresponding FTHUE rehabilitation program based on the objective assessment results of the FTHUE scale, which can enable the development of individualized training programs and effectively integrate the principles of rehabilitation training. In addition, systems developed using markerless sensing technology with high accuracy (Microsoft Azure Kinect and Leap Motion Controller devices) are inexpensive, portable, and suitable for community or home scenarios. This paper reviews the current research progress of VR training devices and intelligent assessment systems for upper limbs in stroke and the limitations of existing devices. It proposes a design concept to achieve a close connection between intellectual assessment and treatment of upper limb motor function in poststroke patients based on the FTHUE scale while integrating assessment and treatment of upper limbs and hands. It aims to provide reference ideas for developing intelligent devices for upper limb rehabilitation.
