Developing a low cost, highly compact holographic imaging based microfluidic cell sorting system using 3D printing

摘要 随着机器视觉和深度学习的进步，基于成像的微流控技术已经展示了它们的 有可能成为各种具有挑战性的医疗应用的精确单细胞分析和分选设备 包括生物打印和细胞图案化，从血液中分离稀有细胞(例如，循环中的肿瘤细胞、镰状细胞) 用于疾病(如癌症、镰刀病)的诊断等。然而，当前基于成像的微流控细胞 分拣系统存在吞吐量低和世界到芯片接口的问题。这些问题很大程度上是 与常规显微成像的景深浅有关，这显著限制了数字 每幅图像可以分析的细胞的数量增加了制造细胞分选的复杂性和成本 集成了成像和阀门驱动功能的芯片。这些问题限制了此类措施的实施 微流控系统在临床诊断和治疗中的应用，特别是在靶细胞极其罕见的情况下。 样本。虽然已经开发了不同类型的高通量细胞分选微流控系统，但它们 要么排序精度不足，要么需要集成不同的排序方法，这进一步 增加了系统的复杂性，降低了系统的可靠性。 这个探索性的R21项目旨在开发一种高通量、低成本和紧凑的成像解决方案- 基于细胞分选的微流控设备，以提高其在关键临床应用中的吸引力。我们的解决方案利用 3D全息成像克服了常规显微成像的景深问题，增加了 细胞检测的特异性，并使用微流控技术实现高通量和高精度的细胞分选。我们 将采用多材料3D打印技术，快速生成反应灵敏的分拣阀，并实现小型化 性能超过文献中的全息成像传感器。我们的方法不会 只有大幅降低成本、时间，并提高制造这些微流体的自动化程度 设备，但也实现了光滑和紧凑的微流体设计，无污染的制造，以及卓越的 在手术时间内保持一致的成像质量，这对于许多临床手术来说是必不可少的。3D打印 本项目所开发的方法可以为低成本、高效率地制造宽幅板提供依据 在医学研究和临床应用中使用的一系列微流控设备(例如，芯片上实验室诊断， 医疗点系统、芯片上器官复制和生物检测)。将3D成像功能与 3D打印将使高通量、高精度和高特异性微流控系统的新设计成为可能 具有在医疗领域中使用的常规微流体所不具备的多种功能。特别是， 在该项目中制造的细胞分选装置可以显著提高基于细胞的液体活检的速度 用于癌症诊断和个性化癌症治疗。

SUMMARY With advancement in machine vision and deep learning, imaging-based microfluidics have demonstrated their potential to serve as precise single cell analysis and sorting devices for various challenging medical applications including bioprinting and cell patterning, sorting rare cells (e.g., circulating tumor cells, sickle cells) from blood for disease (e.g., cancer, sickle diseases) diagnosis, etc. However, current imaging-based microfluidic cell sorting systems suffer from the issues of low throughput and world-to-chip interfacing. These issues are largely related to the shallow depth of field of conventional microscopic imaging, which significantly restricts the number of cells that can be analyzed per image and increases the complexity and cost involved in fabricating cell sorting chips integrated with imaging and valve actuation capabilities. These issues limit the implementation of such microfluidic systems in clinical diagnostics and treatment, particularly in which target cells are extremely rare in samples. Although different types of high throughput cell sorting microfluidic systems have been developed, they either have insufficient sorting precision or require the integration of different sorting methods, which further increases the system complexity and lowers its reliability. This exploratory R21 project aims to develop a high throughput, low cost and compact solution for imaging- based cell sorting microfluidic devices to improve their appeal for critical clinical applications. Our solution utilizes 3D holographic imaging to overcome the depth of field issue of conventional microscopic imaging, increase the specificity of cell detection, and enables high throughput and high precision cell sorting using microfluidics. We will use multi-material 3D printing technique to generate fast responsive sorting valves and miniaturized holographic imaging sensors with performance that exceeds the ones in the literature. Our approach will not only substantially reduce the cost, time, and enhance the degree of automation for fabricating these microfluidic devices, but also enable sleek and compact microfluidic design, contamination-free fabrication, and superior and consistent imaging quality during hours of operation that is essential for many clinical operations. The 3D printing approach developed in this project can provide the basis for low cost and high efficiency fabrication of a broad range of microfluidic devices used in medical research and clinical applications (e.g., lab-on-a-chip diagnostics, point-of-care systems, organ replication-on-a-chip, and bioassays). The integration of 3D imaging capability with 3D printing will enable new designs of high throughput, high precision, and high specificity microfluidic systems with versatile functionalities that are not available in conventional microfluidics used in medical field. In particular, the cell sorting devices fabricated in this proposed project can significantly speed up the cell-based liquid biopsy for cancer diagnostics and personalized cancer treatment.