A placenta-based strategy for improved clinical prediction of fetal growth trajectory using automated image analysis of placental morphology and vascularity

基于胎盘的策略，使用胎盘形态和血管分布的自动图像分析来改进胎儿生长轨迹的临床预测

• 批准号: 10512601
• 负责人: Ipek Oguz
• 金额: $ 60.42万
• 依托单位: VANDERBILT UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-05 至 2027-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10512601
• 关键词: 3D ultrasound; Acoustics; Algorithmic Software; Area; Assessment tool; Award; Biological Markers; Biological Models; Blood; Blood Vessels; Characteristics; Chorion; Clinical; Collaborations; Computers; Custom; Detection; Evaluation; Fetal Growth; Fetal Growth Retardation; Fetal Monitoring; Fetal Weight; Fetus; First Pregnancy Trimester; Gestational Age; Growth; Growth Disorders; Health; Histopathology; Hormonal; Human; Image; Image Analysis; Image Enhancement; Immune; Individual; Infant; Knowledge; Laws; Link; Manuals; Measurement; Measures; Medical Imaging; Metabolic; Methodology; Methods; Modeling; Modernization; Monitor; Morphology; National Institute of Child Health and Human Development; Nutritional; Outcome; PGF gene; Pathologic; Pathology; Pattern; Perinatal; Personal Satisfaction; Physicians; Physics; Placenta; Placental Extracts; Placentation; Play; Postpartum Period; Predictive Factor; Predictive Value; Pregnancy; Prenatal care; Provider; Research; Research Personnel; Risk; Risk Factors; Role; Scientist; Screening procedure; Second Pregnancy Trimester; Serum; Shapes; Site; Small for Gestational Age Infant; Structure; Surface; Techniques; Testing; Time; Ultrasonography; antenatal; antenatal care; automated image analysis; automated segmentation; base; clinical care; deep learning algorithm; density; dynamic system; fetal; healthy pregnancy; histopathological examination; image processing; improved; improved outcome; infant outcome; innovation; interdisciplinary approach; mathematical model; morphometry; novel; novel strategies; open innovation; personalized management; placental morphology; precision medicine; prenatal; prevent; recruit; screening; stillbirth; tool

PROJECT SUMMARY/ABSTRACT Fetal growth restriction is associated with a profound increase in perinatal and even long-term health risk. Antenatal care is key to optimizing outcomes and preventing stillbirth, yet up to half of growth-restricted infants are not identified during pregnancy. The placenta serves a central in maintaining a healthy pregnancy and supporting fetal growth; yet, direct assessment of placental development is glaringly absent from clinical care as there are no practical tools that enable providers to monitor placental development. In recent years, 3D ultrasound (3DUS) has allowed investigators to identify important associations between placental morphology and clinical outcomes using a variety of offline medical image analysis techniques. However, these techniques typically require extensive manual input. Moreover, we have recently developed an innovative tool based on a dynamic model of fetal-placental growth that considers placental growth in the evaluation of fetal growth and can help identify pregnancies at increased risk of growth restriction. However, this tool requires placental volume assessment, which, as mentioned above, remains impractical for clinical use. In this proposal, we will expand and enhance our automated segmentation tools to enable bedside volumetric assessment of the placenta throughout pregnancy. In addition, we will develop novel tools and parameters for assessing placental shape, gross morphology, and vascularity in an effort to identify additional features of placental development that can augment our understanding of placental development and create additional markers of placental health. Taken together, the current proposal leverages an ongoing collaboration between computer scientists and physician-scientists to utilize modern fully automated image analysis methodology to create clinically impactful placental assessment tools that can be integrated into the clinical workflow. The proposed research will allow bedside assessment of placental morphology and vascularity, which can be leveraged into precision medicine approaches and allow for more accurate and reliable surveillance of fetal growth and well-being. Specifically, we will build: 1) Refine and validate a fetal-placental growth model using automated early placental volume and placental histopathology, 2) Extend to include later gestational ages and expand the toolkit to include novel measures of placental shape and vascularity, and 3) create an augmented version of the dynamic model that incorporates the added functionality of our segmentation pipeline, as well as serum biomarkers, to result in a clinically useful tool for monitoring fetal growth. We anticipate that this proposal will significantly change clinical care and create a new, placenta-based paradigm for understanding and managing fetal growth disorders.

项目概要/摘要 胎儿生长受限与围产期甚至长期健康风险的大幅增加有关。 产前护理是优化结局和预防死产的关键，但多达一半的婴儿生长受限 怀孕期间无法识别。胎盘是维持健康妊娠的核心， 支持胎儿生长；然而，临床护理中明显缺乏对胎盘发育的直接评估，因为 没有实用的工具可以让提供者监测胎盘发育。近年来，3D 超声 (3DUS) 使研究人员能够识别胎盘形态之间的重要关联 使用各种离线医学图像分析技术的临床结果。然而，这些技术 通常需要大量的手动输入。此外，我们最近开发了一种基于 胎儿胎盘生长动态模型，在评估胎儿生长时考虑胎盘生长，可以 帮助识别生长受限风险增加的妊娠。然而，该工具需要胎盘体积 评估，如上所述，对于临床使用仍然不切实际。 在本提案中，我们将扩展和增强我们的自动分割工具，以实现床边体积测量 整个怀孕期间胎盘的评估。此外，我们将开发新颖的工具和参数 评估胎盘形状、大体形态和血管分布，以确定胎盘的其他特征 胎盘发育可以增强我们对胎盘发育的理解并创造更多 胎盘健康的标志。 总而言之，当前的提案利用了计算机科学家和 医师科学家利用现代全自动图像分析方法来创建具有临床影响力的 可以集成到临床工作流程中的胎盘评估工具。拟议的研究将允许 胎盘形态和血管分布的床边评估，可用于精准医疗 方法并允许更准确和可靠地监测胎儿生长和健康。具体来说，我们 将构建：1）使用自动化的早期胎盘体积来完善和验证胎儿胎盘生长模型， 胎盘组织病理学，2) 扩展到包括较晚的胎龄，并扩展工具包以包括新的 胎盘形状和血管分布的测量，以及 3) 创建动态模型的增强版本 结合了我们的分割管道的附加功能以及血清生物标志物，从而产生 监测胎儿生长的临床有用工具。 我们预计该提案将显着改变临床护理并创建基于胎盘的新范例 了解和管理胎儿生长障碍。