3D-FACS: 3D image-based fluorescence activated cell sorting

总结 本研究的主要目的是展示一种高通量流式细胞仪系统， 基于高内容的3D细胞图像特征对细胞进行分类。对于在微流体通道中流动的每个单细胞， 该系统将从空间分辨的荧光和散射信号以 1000个细胞/秒。每个多参数三维细胞图像将被重建，数百个图像特征将被重建。 提取，并对空间特征满足用户定义标准的细胞进行排序（3D图像引导 细胞分选）。本质上，所提出的系统结合了高通量细胞分析和分选的优点 荧光激活细胞分选仪（FACS）的功能与高内涵的3D成像显微镜，以提供 研究人员和临床医生前所未有的功能和能力，分析，分类和分离细胞， 单细胞分辨率。该工具的发明有望极大地改变细胞表型研究， 加速细胞类型的发现，并加强对高度异质性生物样品的研究， 肿瘤和大脑 为了实现这一宏伟目标，我们将采取几种创新方法。生产高质量的3D细胞 为了获得在流动通道中快速行进的单个细胞的图像，我们发明了一种无相机成像系统， 设计了一种扫描结构光激发与共焦检测相结合的方案， 将3D空间信息转换为高速光电倍增管（PMT）输出端的时间信号。 对于细胞分选机制，我们采用具有片上压电的微流控芯片/盒设计 致动器来分选细胞，而不会引起可能破坏通过光学成像的细胞的成像的流动抖动 区实现实时图像处理和图像特征提取，以及处理图像的传输和 为了存储大量的3D细胞图像数据，我们提出了一种电子系统， 可编程门阵列（FPGA）模块和图形处理单元（GPU），具有FPGA处理PMT 信号，细胞检测，分割和图像重建，以及分选决策控制， 使GPU提取数百个3D图像相关特征并定义排序标准（即，3D图像- 引导选通）并行。为了评估系统的性能，我们将进行实验来分选细胞 基于蛋白质转运和运输的性质，斑点计数，细胞器跟踪和特征 帮助理解疾病生物学和药物开发。 该仪器将为生物医学界提供一个强大的工具，以推进表型研究， 细胞类型发现，并将基因表达研究与单细胞的细胞表型特征联系起来， 分辨率和高吞吐量。这项研究的影响将是重大而深远的。

SUMMARY The primary goal of the proposed research is to demonstrate a high throughput flow cytometer system that can sort cells based on high-content 3D cell image features. For each single cell flowing in a microfluidic channel, the system will produce cell tomography from spatially resolved fluorescent and scattering signals at a rate of 1000 cells/s. Each multi-parameter 3D cell image will be reconstructed, hundreds of image features will be extracted, and cells with their spatial features meeting the user-defined criteria will be sorted (3D image-guided cell sorting). Essentially the proposed system combines the merits of high throughput cell analysis and sorting capabilities of a fluorescence-activated cell sorter (FACS) with a high-content 3D imaging microscope to offer researchers and clinicians unprecedented features and capabilities to analyze, classify, and isolate cells at single cell resolution. The invention of this tool is anticipated to transform cell phenotype studies, greatly accelerate cell type discoveries, and enhance studies of highly heterogeneous biological samples such as tumors and brain. To realize such ambitious goal, we will take several innovative approaches. To produce high-quality 3D cell images for individual cells travelling fast in a flow channel, we invent a camera-less imaging system using a design that combines scanning structured light excitation and the scheme of confocal detection, which transforms 3D spatial information into temporal signals at the output of high-speed photomultiplier tubes (PMTs). For cell sorting mechanism, we adopt a microfluidic chip/cartridge design with an on-chip piezoelectric actuator to sort cells without causing flow jitters that can disrupt imaging of cells passing the optical imaging area. To achieve real-time image processing and image feature extraction, as well as handling the transport and storage of the large amount of 3D cell image data, we propose an electronic system consisting of a field programmable gate array (FPGA) module and graphics processing unit (GPU), having the FPGA process PMT signals, cell detection, segmentation and image reconstruction, and sorting decision control while having the GPU extract hundreds of 3D image related features and define sorting criteria (i.e. 3D image- guided gating) in parallel. To evaluate the performance of the system, we will perform experiments to sort cells based on the properties of protein translocation and trafficking, spot counting, organelle tracking, and features that help understand the disease biology and drug development. The proposed instrument will offer biomedical community a powerful tool to advance phenotype studies and cell type discoveries, and to link gene expression studies to cell phenotypic characteristics at single cell resolution and high throughput. The impact of the research will be significant and profound.