AI based system for longitudinal, repeated measure analyses of freely moving C. elegans worms

基于人工智能的系统，用于对自由移动的秀丽隐杆线虫进行纵向、重复测量分析

• 批准号: 10258638
• 负责人: JACOB R GLASER
• 金额: $ 45万
• 依托单位: MICROBRIGHTFIELD, LLC
• 依托单位国家: 美国
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-07-09 至 2022-07-08
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10258638
• 关键词: Acetylcholine; Address; Agar; Aging; Alzheimer' s Disease; Animal Behavior; Animal Model; Appearance; Artificial Intelligence; Behavior; Behavioral; Behavioral Research; Biological Assay; Biotechnology; Brain; Brain Diseases; Caenorhabditis elegans; Classification; Collaborations; Complex; Computer software; Data; Development; Disease; Dopamine; Eukaryota; Feasibility Studies; Glutamates; Goals; Human; Image; Individual; Laboratories; Legal patent; Lighting; Locomotion; Longevity; Longitudinal Studies; Market Research; Massachusetts; Measures; Microscope; Molecular; Motion; Mus; Names; National Institute of Mental Health; Nematoda; Neurodegenerative Disorders; Neurodevelopmental Disorder; Neurosciences; Neurotransmitters; Pathologic; Pharmacologic Substance; Phase; Population Analysis; Preparation; Psyche structure; Rattus; Research; Research Design; Research Personnel; Rodent; Schizophrenia; Schools; Serotonin; Speed; Strategic Planning; System; Technology; Testing; Time; Toxicology; Traumatic Brain Injury; Validation; Visual Fields; Work; analytical method; autism spectrum disorder; base; behavioral study; convolutional neural network; design; digital; drug discovery; fighting; free behavior; gamma-Aminobutyric Acid; high throughput screening; innovation; intelligent algorithm; longitudinal analysis; neural network architecture; neurodevelopment; neuropsychiatric disorder; neurotransmitter release; next generation; novel; novel therapeutics; prevent; prototype; receptor; social; usability

Abstract This project aims to develop WormInvestigator™, a novel, highly innovative system for performing automated, high-throughput and longitudinal studies of the behavior of C. elegans worms freely moving and socially interacting on agar plates (hereafter: "freely moving worms") across multiple time points over extended times (e.g., multiple days) with repeated measures designs. Work in Phase I will focus on demonstrating feasibility of our novel, patent pending, WormRecognizer™ technology – the ability to perform automatic, image-based identification of individual C. elegans worms within a group of freely moving worms ("digital tagging of freely moving worms"). Work in Phase II will focus on creating the full functionality of WormInvestigator for the commercial release. The innovation inherent in WormRecognizer will serve as the basis for enabling a game- changing innovation in the field – the ability to perform high throughput longitudinal, repeated measures design analyses of locomotion and other behavior of freely moving C. elegans worms from discrete, non-continuous video sequences. Compared to study designs that have independent groups repeated measures designs offer more statistical power and the possibility to track an effect over time. Specifically, repeated measures designs for analyzing locomotion and other behavior of freely moving worms will allow researchers to definitively assess the likelihood that a particular behavior is associated with a prior behavior, which is impossible without repeated measures designs or impractical continuous imaging and tracking under constant illumination. WormRecognizer will leverage the Deep Convolutional Neural Network (CNN) architecture to perform automatic identification of the tracks of the same worm in videos of groups of freely moving worms recorded at different time points; encouraging pilot data were generated during preparation of this application. C. elegans is increasingly used as a model organism in research focusing on brain mechanisms underlying complex behaviors and pathological alterations thereof, including research into neurodevelopment, Alzheimer's disease, autism, schizophrenia and traumatic brain injury. Thus, WormInvestigator will enable significant advancements in various mental neuroscience applications that use C. elegans as a model organism. Specifically, the fact that C. elegans express many of the neurotransmitters and associated receptors that are found in higher eukaryotes, including humans, makes C. elegans highly attractive for the (high throughput) screening of next generation therapeutics for mental diseases such as Alzheimer's disease, as well as for disorders that rely on neurotransmitter release modulation such as next generation treatments for schizophrenia. We will perform extensive feasibility studies, product validation and usability studies of WormInvestigator in close collaboration with expert neuroscientists. Market research performed during preparation of this application indicated that WormInvestigator will expand the use of C. elegans as a model organism to many laboratories that do not currently use them. A competing technology is not available. We anticipate the global market size for WormInvestigator to be more than 300 systems.

摘要 该项目旨在开发WormInvestigator™，这是一种新颖的，高度创新的系统，用于执行自动化， 高通量和纵向研究C.秀丽蠕虫自由移动和社会性 在琼脂平板上相互作用（以下称为“自由移动的蠕虫”） (e.g.,多天）与重复测量设计。第一阶段的工作将侧重于证明以下方面的可行性： 我们正在申请专利的新型WormRecognizer™技术-能够执行自动的、基于图像的 个体识别C.在一组自由移动的蠕虫中的秀丽线虫蠕虫（“自由移动的线虫的数字标记”） 移动蠕虫”）。第二阶段的工作将集中在创建WormInvestigator的完整功能， 商业释放WormRecognizer固有的创新将成为实现游戏的基础- 不断变化的创新领域-执行高通量纵向、重复测量设计的能力 分析自由活动的C.从离散的、非连续的 视频序列。与具有独立组的研究设计相比，重复测量设计提供了 更强的统计能力和追踪长期效应的可能性。具体而言，重复测量设计 用于分析自由移动蠕虫的运动和其他行为将使研究人员能够明确评估 特定行为与先前行为相关联的可能性，不重复就不可能 测量设计或在恒定照明下不切实际的连续成像和跟踪。蠕虫识别器 将利用深度卷积神经网络（CNN）架构来执行自动识别 在不同时间点记录的自由移动的蠕虫群的视频中的同一蠕虫的轨迹; 在准备这一申请的过程中，产生了令人鼓舞的试验数据。C.越来越多地被用作 一种模式生物，研究重点是复杂行为和病理的大脑机制。 包括对神经发育、阿尔茨海默病、自闭症、精神分裂症和 创伤性脑损伤因此，WormInvestigator将在各种心理学领域取得重大进展。 神经科学应用程序使用C。elegans作为模式生物。具体来说，事实上，C。埃莱甘斯快报 在高等真核生物，包括人类， 使C. elegans对下一代精神疾病治疗剂的（高通量）筛选非常有吸引力， 阿尔茨海默病等疾病，以及依赖于神经递质释放调节的疾病 比如精神分裂症的下一代治疗方法我们将进行广泛的可行性研究， WormInvestigator的验证和可用性研究与专家神经科学家密切合作。市场 在准备本申请期间进行的研究表明，WormInvestigator将扩大使用 C.秀丽线虫作为模式生物，许多实验室目前没有使用它们。一项竞争技术是 不可用.我们预计WormInvestigator的全球市场规模将超过300个系统。