Characterization of the Neurobiological Profiles of Young Adults with and without Developmental Language Disorder (DLD)

摘要 大约7%的学龄儿童患有发展性语言障碍(DLD)，使其成为 儿童语言障碍患病率最高。DLD将个人置于学业失败的风险中，社会 与世隔绝、焦虑、抑郁、情绪调节不良、青少年监禁(65%)、反复犯罪 (70%)。DLD一直持续到成年期，保守估计表明，美国有1200万成年人 各州有DLD，但因为行为表型可以与典型个人的行为表型重叠，所以它们可能 不再有资格获得支持服务。值得注意的是，尽管这些患有DLD的年轻人看起来很正常，但他们的 语言能力与不同于典型个体的大脑结构和皮质动力学有关， 提示DLD的神经特征可能是该病的一个重要标志。新的机器学习方法 使神经发育障碍的神经科学发生了革命性的变化，并在记录光学 功能近红外光谱(FNIRS)信号与神经解剖学数据的结合，使得捕捉脑功能的空间信息成为可能。 口语处理的时间动力学对于语音、语言和听力来说是可行且经济有效的 人口。这两个领域的专业知识对于高影响力的演讲语言研究至关重要。候选人 是一名知名的调查员，拥有超过30年的研究和外部资助的良好记录 在DLD地区工作了几年。增强的目标是增强应聘者目前在DLD方面的专业知识 通过接受fNIRS神经成像和更新的计算建模技术方面的高级培训，以保持 候选人的研究计划与新兴和不断发展的神经成像和计算建模同步进行 接近了。增强的目标是：(1)提高应聘者在尖端fNIR方面的技能 方法，(2)将计算建模纳入候选人的研究计划，以及(3)催化新的 与认知神经学家、光学成像仪和计算模型师的研究合作。说教 应用机器学习和计算建模的课程作业(学期长的课程)，1：1会议， 学术旅行和小规模研究项目将为以下人员提供增强的体验 增强候选人的研究技能，与其他领域的科学家开展新的合作 与DLD相关。研究项目将为应聘者提供实践机会以获得 在具有光学专业知识的年轻、有前途的导师组成的卓越团队的指导下获得专业知识 成像(FNIRS)，以及功能和结构脑成像(fMRI，MRI)，以及一组资深合作者 拥有计算建模和认知神经科学方面的专业知识。一群18岁的年轻人；0-21岁；0 有/没有DLD(n=44)将完成标准化评估、结构MRI和一系列fNIR 任务，然后将其用作输入以推导出DLD的多维模型。总体而言，增强 将极大地增强知名DLD研究人员的当前研究轨迹，并提供 美国国立卫生研究院新基金，用于开发DLD的多维神经生物学衍生模型。

ABSTRACT Approximately 7% of school-aged children have Developmental Language Disorder (DLD), making it one of the highest prevalence of the child language disorders. DLD places individuals at risk for academic failure, social isolation, anxiety, depression, poor emotional regulation, juvenile incarceration (65%), and repeat offending (70%). DLD persists into adulthood, with conservative estimates indicating that 12 million adults in the United States have DLD, but because the behavioral phenotype can overlap with that of typical individuals', they may no longer qualify for support services. Notably, although these young adults with DLD "appear" normal, their language abilities are linked to brain structure and cortical dynamics that differ qualitative from typical individuals, suggesting that neural signature of DLD may be a critical marker of the disorder. New machine learning methods have revolutionized the neuroscience of neurodevelopmental disorders and advances in registering the optical signal of functional near infrared spectroscopy (fNIRS) to neuroanatomical data now make capturing the spatial- temporal dynamics of spoken language processing feasible and cost-effective for speech, language, and hearing populations. Expertise in these two domains is critical for high impact speech-language research. The candidate is an established investigator with a strong record of research and extramural funding spanning more than 30 years in the area of DLD. The goal of the enhancement is to augment the candidate's current expertise in DLD by gaining advanced training in fNIRS neuroimaging and newer computational modeling techniques to keep the candidate's program of research in-step with emerging and evolving neuroimaging and computational modeling approaches. The goals of the enhancement are to: (1) advance the candidate's skills in cutting-edge fNIRS methods, (2) incorporate computational modeling into the candidate's program of research, and (3) catalyze new research collaborations with cognitive neuroscientists, optical imagers, and computational modelers. Didactic course work in applied machine learning and computational modeling (semester long courses), 1:1 meetings, scholarly travel and a small-scale research project will provide the enhanced experience to substantially augment the research skills of the candidate, seed new collaborations with scientists in other fields whose work is relevant to DLD. The research project will provide the hands-on opportunity for the candidate to acquire expertise under the mentorship of a superior team of young, up and coming mentors with expertise in optical imaging (fNIRS), and functional and structural brain imaging (fMRI, MRI), and a group of senior collaborators with expertise in computational modeling and cognitive neuroscience. A group of young adults ages 18;0 - 21;0 with/without DLD (n = 44) will complete standardized assessments, a structural MRI, and a series of fNIRS tasks, which will then be used as inputs to derive multidimensional models of DLD. Overall, the enhancement will significantly augment the current research trajectory of a well-established DLD researcher and provide the foundation for new NIH funding to develop multidimensional neurobiologically derived models of DLD.