论文标题
灵活的虚拟现实体系,用于神经康复和生活质量改善
Flexible Virtual Reality System for Neurorehabilitation and Quality of Life Improvement
论文作者
论文摘要
随着预期寿命的大部分增加,许多神经系统疾病的发病率也在不断增长。为了改善受神经系统障碍影响的身体功能,必须定期进行康复程序,并且必须定期进行。不幸的是,神经康复程序在成本,可及性和缺乏治疗师方面存在缺点。本文使用虚拟现实(INREX-VR)提出了沉浸式神经居住练习,这是我们使用虚拟现实创新的沉浸式神经居住系统。该系统基于彻底的研究方法,能够捕获实时用户运动并评估上肢和下肢的关节活动,记录训练会议并保存肌电图数据。第一人称视角的使用增加了浸入式,并且在骨骼钻机上应用的HTC VIVE系统和逆运动学原理的帮助下,运动的关节范围是计算的。虚拟治疗师证明了教程练习,因为它们是与现实生活中的医生一起记录的,并且可以通过远程医学来监视和配置会议。在游戏设置中实践复杂的运动,鼓励自我完善和竞争。最后,我们提出了一个培训计划和初步测试,该测试在准确性和用户反馈方面显示出令人鼓舞的结果。作为未来的发展,我们计划提高系统的准确性,并根据神经网络进行无线替代方案。
As life expectancy is mostly increasing, the incidence of many neurological disorders is also constantly growing. For improving the physical functions affected by a neurological disorder, rehabilitation procedures are mandatory, and they must be performed regularly. Unfortunately, neurorehabilitation procedures have disadvantages in terms of costs, accessibility and a lack of therapists. This paper presents Immersive Neurorehabilitation Exercises Using Virtual Reality (INREX-VR), our innovative immersive neurorehabilitation system using virtual reality. The system is based on a thorough research methodology and is able to capture real-time user movements and evaluate joint mobility for both upper and lower limbs, record training sessions and save electromyography data. The use of the first-person perspective increases immersion, and the joint range of motion is calculated with the help of both the HTC Vive system and inverse kinematics principles applied on skeleton rigs. Tutorial exercises are demonstrated by a virtual therapist, as they were recorded with real-life physicians, and sessions can be monitored and configured through tele-medicine. Complex movements are practiced in gamified settings, encouraging self-improvement and competition. Finally, we proposed a training plan and preliminary tests which show promising results in terms of accuracy and user feedback. As future developments, we plan to improve the system's accuracy and investigate a wireless alternative based on neural networks.