EAGER: An Exploratory Pilot Project to Build an Intelligent Human-Computer Interface System for Neurological Disorder Assessment and Rehabilitation

在这次热切期待中，PI将探讨与开发和使用游戏数据以评估和康复神经系统疾病有关的问题。重点是脑瘫（CP），一种导致各种运动和其他损伤的疾病。CP最常见的症状是肌肉不协调、肌肉僵硬、反射过度和步态受损，治疗方法包括物理治疗（PT）和职业治疗（OT）、语言治疗、药物、手术和矫形器。最初，PI将研究基于游戏的上肢系统。传统上，OT专家使用主观判断结合手工能力分类系统（MACS）来衡量运动技能，而与皮质水平的活动无关。PI的方法是将皮质活动与运动活动联系起来。她将探索计算机游戏和可穿戴传感器的耦合和集成，用于神经和运动评估测试，以及家庭长期康复。特别是，她将设计一系列电脑游戏来对应于OT练习，并为具有运动技能评估的CP类型建立一套初始特征分类器。她还将为远程康复和周期性评价方法建立计算机基础设施。本研究的主要成果将是远程康复的初始设置，该设置使用机器强化学习并融合从各种传感器收集的多模态信息。将电脑游戏与脑成像和传统的康复结果相结合，是一种激进的、未经检验的、但可能改变慢性疾病（如cp）医疗保健实践的方法。PI的长期目标是开发一个名为CPLAY的智能系统，其前端是一组电脑游戏，为患有CP的儿童提供受控刺激，以促进所需的运动反应，并生成用于诊断和康复治疗的性能数据。CPLAY的后端将是一组计算引擎，以实现数据记录（来自玩游戏的孩子）、数据融合、分析和决策支持。实验室版本的CPLAY将用于功能性近红外（fNIR）成像，而家庭版本将用于康复，在游戏过程中使用各种额外的传感器捕获和融合数据。fNIR成像用于确定神经可塑性和运动恢复。@home版本会随时间跟踪表现，考虑游戏环境，如果需要还可以远程监控。这两个版本都将允许游戏调整，以提供个性化的治疗。CPLAY方法有望降低成本，促进家庭参与康复。项目成果将包括广泛适用于其他神经系统疾病的范例、方法和工具。

In this EAGER the PI will explore issues relating to the development and use of data from game playing for assessment and rehabilitation of neurological disorders. The focus is on Cerebral Palsy (CP), a disease that causes a variety of motor and other impairments. The most common symptoms of CP are a lack of muscle coordination, stiff muscles, exaggerated reflexes, and impaired gait, and treatment includes Physical Therapy (PT) and Occupational Therapy (OT), speech therapy, drugs, surgery, and orthotic devices. Initially, the PI will investigate game-based systems for the upper extremities. Traditionally, OT experts use subjective judgment in conjunction with the Manual Ability Classification System (MACS) to measure the motor skills without connection to cortical level activity. The PI's approach is to connect cortical activity with motor activity. She will explore the coupling and integration of computer games and wearable sensors for both neurological and motor assessment testing, as well as for long-term rehabilitation at home. In particular, she will design a family of computer games that correspond to OT exercises and build an initial set of feature classifiers for types of CP with motor skill assessment. She will also build computer infrastructure for remote rehabilitation and a cyclical evaluation methodology. The primary outcome of this research will be an initial setup for remote rehabilitation that uses machine reinforcement learning and fuses multimodal information collected from a variety of sensors. Using computer games in conjunction with brain imaging and traditional rehabilitation outcomes is a radical and untested but potentially transformative approach to healthcare practices for chronic conditions such as CP.Broader Impacts: The PI's long-term goal is to develop an intelligent system called CPLAY, whose front end is a set of computer games to provide controlled stimuli to children with CP in order to facilitate a desired motor response and generate performance data for diagnosis and rehabilitation treatment. CPLAY's backend will be a set of computational engines to enable data logging (from a child playing the game), data fusion, analysis and decision support. The @lab version of CPLAY will be used for functional near infrared (fNIR) imaging, while the @home version will be used for rehabilitation with various additional sensors capturing and fusing data during game playing. The fNIR imaging is used to determine neuro-plasticity and motor recovery. The @home version tracks performance over time, considers context of the game, and can be monitored remotely if needed. Both versions will allow for game adjustments to provide personalized treatment. The CPLAY approach promises to lower costs and facilitate family engagement in the rehabilitation. Project outcomes will include a paradigm, methodology and tools with broad applicability to other neurological disorders.