The ability to rapidly assay morphological and intracellular molecular variations within large heterogeneous populations of cells is essential for understanding and exploiting cellular heterogeneity. Optofluidic time-stretch microscopy is a powerful method for meeting this goal, as it enables high-throughput imaging flow cytometry for large-scale single-cell analysis of various cell types ranging from human blood to algae, enabling a unique class of biological, medical, pharmaceutical, and green energy applications. Here, we describe how to perform high-throughput imaging flow cytometry by optofluidic time-stretch microscopy. Specifically, this protocol provides step-by-step instructions on how to build an optical time-stretch microscope and a cell-focusing microfluidic device for optofluidic time-stretch microscopy, use it for high-throughput single-cell image acquisition with sub-micrometer resolution at >10,000 cells per s, conduct image construction and enhancement, perform image analysis for large-scale single-cell analysis, and use computational tools such as compressive sensing and machine learning for handling the cellular 'big data'. Assuming all components are readily available, a research team of three to four members with an intermediate level of experience with optics, electronics, microfluidics, digital signal processing, and sample preparation can complete this protocol in a time frame of 1 month.

No abstract

In the last decade, high-content screening based on multivariate single-cell imaging has been proven effective in drug discovery to evaluate drug-induced phenotypic variations. Unfortunately, this method inherently requires fluorescent labeling which has several drawbacks. Here we present a label-free method for evaluating cellular drug responses only by high-throughput bright-field imaging with the aid of machine learning algorithms. Specifically, we performed high-throughput bright-field imaging of numerous drug-treated and -untreated cells (N = ~240,000) by optofluidic time-stretch microscopy with high throughput up to 10,000 cells/s and applied machine learning to the cell images to identify their morphological variations which are too subtle for human eyes to detect. Consequently, we achieved a high accuracy of 92% in distinguishing drug-treated and -untreated cells without the need for labeling. Furthermore, we also demonstrated that dose-dependent, drug-induced morphological change from different experiments can be inferred from the classification accuracy of a single classification model. Our work lays the groundwork for label-free drug screening in pharmaceutical science and industry.

The drug susceptibility of leukemia cells in whole blood is evaluated by using extreme-throughput imaging flow cytometry with deep learning.