Robot child interactions as an intervention tool for children with autism

描述（由申请人提供）：自闭症谱系障碍（ASD）的发病率不断上升，这对研究人员开发新的ASD儿童干预技术提出了挑战。然而，大多数关于ASD干预的研究都集中在传统和现代的应用行为分析方法上，这些方法针对儿童的语言、认知和学前技能。这些方法通常要求自闭症儿童每周进行30-40小时的一对一强化干预。相比之下，由于各种原因，管理机器人儿童互动作为一种干预技术可能是一项重大创新。机器人可以作为临床医生的辅助工具，以某种程度的标准化管理评估或治疗方案的组成部分，并可能协助数据收集过程。机器人的可编程性使我们能够系统地增加治疗过程中社会互动的复杂性。证据还表明，自闭症儿童喜欢与机器人互动，因为它们的行为简单且可预测。最重要的是，机器人是具身的存在，可以鼓励自闭症儿童进行全身互动，包括模仿、轮流和共同注意。目的：本研究的总体目标是为诊断为ASD的儿童开发新的干预技术。具体而言，建议项目将审查

DESCRIPTION (provided by applicant): The rising incidence of Autism Spectrum Disorders (ASD) challenges researchers to develop novel intervention techniques for children with ASD. Nevertheless, the majority of the research on ASD interventions is focused on traditional and contemporary Applied Behavioral Analysis approaches that target a child's language, cognitive, and preacademic skills. These approaches typically require a child with ASD to engage in intensive 1:1 intervention for 30-40 hours per week. In contrast, administering robot child interactions as an intervention technique may be a significant innovation for various reasons. A robot could serve as an adjunct to a clinician to administer components of an assessment or treatment protocol with some level of standardization and may assist in the data collection process. The programmability of a robot allows us to systematically increase the complexity of social interactions within treatment. Evidence also suggests that children with ASD enjoy interacting with robots due to their simple and predictable behaviors. Most of all, robots are embodied beings that encourage a child with ASD to engage in whole body interactions involving imitation, turn taking, and joint attention bids. Aims: The overall goal of this research is to develop novel intervention techniques for children diagnosed with ASD. Specifically, the proposed project will examine the efficacy of robot child interactions for enhancing social, communication, and perceptuo-motor skills of children with high and low functioning ASD. Our preliminary data on typically developing (TD) children who underwent short-term training with a humanoid robot suggested that all children enjoyed interacting with the robot. The young children performed behaviors such as looking at the robot, greeting the robot, and talking to the robot. They also imitated the complex limb movements performed by the robot. In terms of non-verbal communication, they engaged in turn taking and joint attention behaviors with their teachers and peers. In the proposed project, we have two phases of study. In the first R21 phase we examine robot child interactions in three different projects: a) interactions of TD children between 4 and 8 years of age during short-term training, b) interactions of the same group during interpersonal synchrony tasks, c) Robotic systems development and d) a 5-week intervention for children with ASD using a single subject design. In the R33 phase, we conduct a group design study in which 12 children with ASD between 4 and 8 years of age will receive a 10-week robot child interaction protocol and be compared to a control group of children with ASD. Pre- and post-training assessments will include the PDD- Behavioral Inventory, Movement-ABC, and the Florida Apraxia Battery. We expect the children in the training group to demonstrate advances general social communication measures as well as specific improvements in imitation, nonverbal communication, as well as motor coordination during robot-child interactions. Overall, robot based social intervention is an uncharted, promising intervention tool for enhancing social, communication and motor skills in children with ASD.