Volume-based Analysis of 6-month Infant Brain MRI for Autism Biomarker Identification and Early Diagnosis

项目总结/摘要 标题：6个月婴儿脑MRI的基于体积的分析，用于自闭症生物标志物识别和 早期诊断 自闭症谱系障碍（ASD）是一种复杂的发育障碍，其特征在于社交能力的缺陷。 互动，语言技能，重复的刻板行为和有限的兴趣。基于一种新 政府调查显示，每45名儿童（3至17岁）中就有1名被诊断患有ASD， 疾病控制和预防中心先前估计的2011-2013年患病率为1/68。 基于体积的神经成像数据分析在成人自闭症研究中发挥着越来越重要的作用， 并发现了广泛的结构和功能异常。然而，现有的基于数量的分析 为成人大脑开发的工具不适合婴儿研究，因为在脑组织中存在巨大挑战 分割和ROI标记，由极低的组织对比度引起。 要成为婴儿神经影像学研究的独立调查员，候选人在此K 01中提出 申请接受临床现象学和儿童发展认知神经科学培训， 自闭症儿童，发育神经生物学和神经发育障碍，以及生物统计学。这些 培训活动将极大地增强候选人在ASD，婴儿神经成像映射和 为他的长期目标奠定了坚实的基础，成为开发基于成像的 自闭症的早期生物学标记 在研究计划中，候选人将创建一套独特的婴儿特定的，基于体积的 神经成像分析工具，使早期大脑发育的准确表征， 自闭症婴儿，以及提高能力，在早期识别生物标志物和早期 诊断高危婴儿。具体来说，一种统一的颅骨剥离和组织分割的新方法将 发展（目标1）。此外，一个新的地图集引导的多通道森林学习将被提出用于ROI标记 (Aim 2）。通过准确的组织分割和ROI标记，将可以进行基于ROI的体积测量。 进行并用于识别自闭症风险的早期指标或生物标志物（目标3）。最后，早期诊断 将对婴儿进行（目标4）。这项研究的结果将有助于确定早期生物标志物的风险， 自闭症，并设计有针对性的先发制人的干预策略。所有创建的工具和地图集将 集成并免费向公众发布，例如通过NITRC（www.nitrc.org）。

Project Summary/Abstract Title: Volume-based analysis of 6-month infant brain MRI for autism biomarker identification and early diagnosis Autism spectrum disorder (ASD) is a complex developmental disability, characterized by deficits in social interaction, language skills, repetitive stereotyped behaviors, and restricted interests. Based on a new government survey, it shows 1 in 45 children (ages 3 to 17) are diagnosed with ASD, a significant increase from Centers for Disease Control and Prevention's previously estimated prevalence of 1 in 68 from 2011-2013. Volume-based analysis of neuroimaging data is playing an increasingly critical role in adult autism studies, and has revealed widespread structural and functional abnormalities. However, existing volume-based analysis tools developed for adult brains are ill-suited for infant studies, due to great challenges in brain tissue segmentation and ROI labeling, caused by the extremely low tissue contrast. To become an independent investigator on infant neuroimaging research, the candidate proposes in this K01 application to receive training in clinical phenomenology and child developmental cognitive neuroscience of children with ASD, developmental neurobiology and neurodevelopmental disorders, and biostatistics. These training activities will greatly augment the candidate's background in ASD, infant neuroimaging mapping and establish a solid foundation for his long-term goal of being a leading researcher on developing imaging-based early biological markers for autism. In the research plan, the candidate will create a unique suite of infant-specific, volume-based neuroimaging analysis tools that enable accurate characterization of early brain development in autistic infants, as well as improved capabilities in early identification of biomarkers and early diagnosis of at-risk infants. Specifically, a new method for unified skull stripping and tissue segmentation will be developed (Aim 1). Also, a new atlas-guided multi-channel forest learning will be proposed for ROI labeling (Aim 2). With the accurate tissue segmentation and ROI labeling, ROI-based volume measurements will be performed and used to identify early indicators or biomarker of risk for autism (Aim 3). Finally, early diagnosis of infants will be performed (Aim 4). Results from this research will help identify early biomarkers of risk for autism and also design targeted preemptive intervention strategies. All created tools and atlases will be integrated and released freely to the public, such as through NITRC (www.nitrc.org).