Three-dimensional fluorescence imaging flow cytometry at up to million frames per second

流式细胞术是高速分析大细胞群的首选工具，但需要权衡缺乏细胞计数的缺点。 胞内空间信息成像流式细胞术（IFC）已经成为一种新的工具， 显微镜的优点与流式细胞术的高速。但是，它们只能提供2D图像， 确定细胞特征的三维（3D）分布，具有有限的视场（FOV），并且需要 精确控制流体系统，以最小化由于不受控制的细胞旋转或平移而导致的图像模糊 穿过视野3D成像的缺乏导致对象位置的模糊和焦深的模糊 这是由于3D细胞投影到2D图像中。尽管在过去的几十年中， 开发了实际获取三维（3D）空间信息的方法， 样本量仍然是他们最大的限制。因此，本提案的目标是制定下一个 新一代3D成像流式细胞仪，具有高通量和高内容能力，可进行3D成像 每秒数百到数千个细胞和球状体，第一次以高分辨率。我们建议 开发这样的细胞计数方法，使用一种新的显微镜方法，线激发阵列检测显微镜 （LEAD），其可以以当前1D细胞仪的速率在大视场中对对象进行成像，但是具有高3D分辨率。 分辨率和高信噪比（SNR）。我们提出的LEAD流式细胞仪是一种快速扫描光片 显微镜能够MHz的帧速率。我们将开发最快的MHz线扫描方法， 由啁啾频率信号驱动的纵向声光偏转器。我们将扫描的光片成像 使用线性硅光电倍增管阵列，其将提供如此快速扫描时所需的灵敏度， 以及这种高帧速率所需的并行读出。首先，我们将开发线性LEAD 3D成像流程 在亚微米尺度分辨率和小FOV下的流式细胞术。尽管我们的初步数据显示， 能够以100 kHz - MHz的帧速率在如此高的分辨率和高信噪比下成像，我们将进行实验 测量SNR以确定LEAD细胞术的操作范围。在第二个目标中，我们将增加 通过开发具有贝塞尔光束的双光子LEAD成像流式细胞术的FOV。为了支持更大的FOV， 我们会利用八个十六通道的数据采集卡，开发一个一百二十八通道的数据采集系统。第三 我们将开发一个最先进的计算基础设施，允许文件传输高达25 GB/s， 存储容量（>100 TB），分析仅需3倍的成像时间。我们将使用2个深度学习模型， 分析.如果成功，这种高风险/高回报的提议将改变成像流式细胞术的前景。 所提出的3D成像流式细胞仪可以在各种生物医学领域提供改进的细胞和球体分析。 癌症生物学、微生物学、免疫学、血液学和干细胞生物学等领域。改善的灵敏度 将帮助用户提高研究成果或诊断具有更高统计能力的患者。

Flow cytometry is the tool of choice for high-speed analysis of large cell populations, with the tradeoff of lacking intracellular spatial information. Imaging flow cytometry (IFC) has emerged as a new tool that combines advantages of microscopy with the high speed of flow cytometry. However, they can only provide 2D images to determine three-dimensional (3D) distribution of cellular features, have a limited field of view (FOV), and require precise control of the fluidic system to minimize image blurring due to uncontrolled cell rotation or translation across the FOV. The absence of 3D imaging results in ambiguity of object locations and blurring by focal depth due to the projection of a 3D cell into a 2D image. Although in the last decades flow cytometry systems that can actually acquire three-dimensional (3D) spatial information were developed, constraints related to resolution and samples size remained as their biggest limitation. Therefore, the goal of this proposal is to develop the next generation 3D imaging flow cytometers with high-throughput and high-content capabilities for 3D imaging of hundreds to thousands of cells and spheroids per second with high resolution, for the first time. We propose to develop such a cytometry method, using a novel microscopy method, Line Excitation Array Detection microscopy (LEAD), that can image objects in large field of views at the rate of current 1D cytometers, but with high 3D resolution and high signal-to-noise ratios (SNR). Our proposed LEAD cytometer is a fast-scanned light-sheet microscope capable of MHz frame rates. We will develop the fastest MHz line-scanning method using a longitudinal acousto-optic deflector driven by a chirped frequency signal. We will image the scanned light sheet using a linear silicon photomultiplier array, which will provide the sensitivity required when scanning so quickly, and the parallel readout required for such high frame rates. First, we will develop linear LEAD 3D imaging flow cytometry at sub-micron scale resolution and small FOVs. Although our preliminary data indicates we will be able to image at 100 kHz – MHz frame rates at such high resolution with high SNR, we will perform experiments measuring the SNR to determine the operating range of LEAD cytometry. In the second aim, we will increase the FOV by developing two-photon LEAD imaging flow cytometry with Bessel beams. To support the larger FOV, we will develop a 128-channel data acquisition system using eight 16-channel data acquisition cards. In the third aim, we will develop a state-of-the-art computational infrastructure that allows for file transfers up to 25 GB/s, storage (>100 TB), and analysis that only takes 3x the imaging time. We will use 2 deep learning models for analysis. If successful, this high-risk/high-reward proposal would alter the imaging flow cytometry landscape. The proposed 3D imaging flow cytometer can offer improved cell and spheroid analysis in diverse biomedical fields such as cancer biology, microbiology, immunology, hematology, and stem cell biology. Improved sensitivity will help users to improve research outcomes or diagnose patients with higher statistical power.