Parallelized Imaging and Automated Analysis of Zebrafish Assays with a Gigapixel Microscope

意义：高通量光学显微镜正在改变遗传学、药物发现等研究领域 和神经科学大规模光学分析现在通常使用数千张高分辨率图像来提供关键见解 进入人体，我们的大脑和影响我们的疾病当今的光学显微镜及其相关图像 然而，处理软件还远未达到理想状态。目前的显微镜无法形成细胞尺度分辨率的图像 这从根本上限制了我们监测这些物体详细运动的能力。 生活系统对于筛选许多体内模型生物体，例如斑马鱼，这种限制阻止了当前的设置， 以细胞分辨率同时观察多个自由移动的生物体，这使得高分辨率的体内筛选成为可能。 实验具有挑战性和耗时。这也导致了缺乏有效的图像处理软件的高- 分辨率生物图像分析。建议：在成功的第一阶段和第二阶段项目的过程中，Ramona Optics 开发了一种新的“微型相机阵列显微镜”（MCAM），通过捕获视频克服了上述限制， 每帧高达10亿像素，可以解析数百个自由游动的生物体，细节接近细胞水平（5 µm/像素）。在这项提案中，Ramona Optics将生产一套软件，充分利用MCAM的 信息丰富的录音。该软件将自动跟踪和测量关键的形态和荧光特征 在一个完整的孔板内的所有斑马鱼，同时，以取代目前繁琐和耗时的任务。我们的MCAM 它的新软件将改变目前依赖于体内生物筛选的毒理学和药理学研究， 在更短的时间内产生更多的实验见解，并且具有比当前方法更高的准确性和可重复性。 SA 1：用于并行化幼虫跟踪和视频配准的MCAM软件：Ramona Optics将生产 Python软件，用于跟踪、裁剪和登记MCAM FOV内的所有幼虫，以生成以幼虫为中心的视频， 随后的形态学分析。我们将展示注册数百个幼虫成像视频的能力 同时（<0.1%的错误率），以将保存的数据减少40倍，并生成标准化的每个生物体数据集。 SA 2：形态学终点的自动注释：与俄勒冈州的Tanguay实验室合作 在北卡罗来纳州州立大学的约德实验室的帮助下，我们将创建算法来自动计算10条斑马鱼幼虫 形态端点，如注视方向，胸鳍位置，身体弯曲，我们将在一系列的验证， 毒理学滴定实验（跨实验室验证），以证明至少100倍的速度，以目前的筛选方法。 SA 3：自动明场视频分析：然后我们将扩展我们的自动形态分析软件， 明视野视频，以提供时间依赖性的每种生物体统计分析。我们将通过跟踪眼睛来展示这个软件 同时在96只幼虫上进行成角试验，以证明视敏度作为毒理学筛选的新终点（Tanguay Lab）。 SA 4：实现高通量荧光视频分析：我们将扩展我们的并行化软件， 在96孔板内的所有幼虫中定位荧光，然后在免疫毒理学实验（Yoder Lab）中对其进行测试。 这个IIB阶段项目的成果将是一个旗舰MCAM软件套件，准备部署到所有MCAM用户。

Significance: High-throughput optical microscopy is currently transforming the research fields of genetics, drug discovery and neuroscience. Large-scale optical assays now routinely use thousands of high-resolution images to offer critical insights into the human body, our brain and the diseases that affect us. Today's optical microscopes and their associated image processing software, however, are still far from ideal. Current microscopes cannot form images with cellular-scale resolution over an area larger than a few square centimeters, which fundamentally limits our ability to monitor the detailed movements of living systems. For screening many in vivo model organisms, such as zebrafish, this limitation prevents current setups from simultaneously observing multiple freely moving organisms at cellular resolution, which makes high-resolution in vivo screening experiments challenging and time-consuming. It has also led to a scarcity of effective image processing software for high- resolution organism image analysis. Proposal: Over the course of successful Phase I and Phase II projects, Ramona Optics has developed a new “micro-camera array microscope” (MCAM) that overcomes the above limitations by capturing video with up to 1 gigapixel per frame, which can resolve hundreds of freely-swimming organisms at near-cellular-level detail (5 µm/pixel). In this proposal, Ramona Optics will produce a suite of software that takes full advantage of the MCAM's information-rich recordings. This software will automatically track and measure key morphological and fluorescence features across all zebrafish within a full well plate, simultaneously, to replace currently tedious and time-consuming tasks. Our MCAM and its new software will transform current toxicology and pharmacology research that relies on in vivo organism screening, by producing more experimental insights in less time and with higher accuracy and repeatability than current methods. SA1: MCAM software for parallelized larval tracking and video registration: Ramona Optics will produce Python software to track, crop and register all larvae within the MCAM FOV to produce per-larvae centered video for subsequent morphological analysis. We will demonstrate the ability to register video of hundreds of larvae imaged simultaneously (<0.1% error rate) to reduce saved data by 40X and produce standardized per-organism datasets. SA2: Automated annotation of morphological endpoints: Working with the Tanguay Lab at Oregon State University and the Yoder Lab at NC State University, we will create algorithms to automatically compute 10 larval zebrafish morphological endpoints, such as gaze direction, pectoral fin position, and body curvature, which we will verify in a series of toxicology titration experiments (validated across labs) to demonstrate at least 100X speed-up to current screening methods. SA3: Automated bright-field video analysis: We will then extend our automated morphology analysis software to bright-field video to provide a time-dependent per-organism statistical analysis. We will showcase this software by tracking eye angle across 96 larvae simultaneously to demonstrate visual acuity as a new endpoint in toxicology screens (Tanguay Lab). SA4: Enabling high-throughput fluorescence video analysis: We will extend our parallelized software to monitor localized fluorescence across all larvae within a 96 well-plate, and then test it in an immunotoxicology experiment (Yoder Lab). The outcome of this Phase IIB project will be a flagship MCAM software suite ready for deployment to all MCAM users.