High-throughput multiparametric imaging flow cytometry: toward diffraction-limited sub-cellular detection and monitoring of sub-cellular processes.

We present a sheathless, microfluidic imaging flow cytometer that incorporates stroboscopic illumination for blur-free fluorescence detection at ultra-high analytical throughput. The imaging platform is capable of multiparametric fluorescence quantification and sub-cellular localization of these structures down to 500 nm with microscopy image quality. We demonstrate the efficacy of the approach through the analysis and localization of P-bodies and stress granules in yeast and human cells using fluorescence and bright-field detection at analytical throughputs in excess of 60,000 and 400,000 cells/s, respectively. Results highlight the utility of our imaging flow cytometer in directly investigating phase-separated compartments within cellular environments and screening rare events at the sub-cellular level for a range of diagnostic applications.

A weakly supervised deep learning approach for label-free imaging flow-cytometry-based blood diagnostics.

The application of machine learning approaches to imaging flow cytometry (IFC) data has the potential to transform the diagnosis of hematological diseases. However, the need for manually labeled single-cell images for machine learning model training has severely limited its clinical application. To address this, we present iCellCnn, a weakly supervised deep learning approach for label-free IFC-based blood diagnostics. We demonstrate the capability of iCellCnn to achieve diagnosis of Sézary syndrome (SS) from patient samples on the basis of bright-field IFC images of T cells obtained after fluorescence-activated cell sorting of human peripheral blood mononuclear cell specimens. With a sample size of four healthy donors and five SS patients, iCellCnn achieved a 100% classification accuracy. As iCellCnn is not restricted to the diagnosis of SS, we expect such weakly supervised approaches to tap the diagnostic potential of IFC by providing automatic data-driven diagnosis of diseases with so-far unknown morphological manifestations.

Deformation of leukaemia cell lines in hyperbolic microchannels: investigating the role of shear and extensional components.

The mechanical properties of cells are of enormous interest in a diverse range of physio and pathological situations of clinical relevance. Unsurprisingly, a variety of microfluidic platforms have been developed in recent years to study the deformability of cells, most commonly employing pure shear or extensional flows, with and without direct contact of the cells with channel walls. Herein, we investigate the effects of shear and extensional flow components on fluid-induced cell deformation by means of three microchannel geometries. In the case of hyperbolic microchannels, cell deformation takes place in a flow with constant extensional rate, under non-zero shear conditions. A sudden expansion at the microchannel terminus allows one to evaluate shape recovery subsequent to deformation. Comparison with other microchannel shapes, that induce either pure shear (straight channel) or pure extensional (cross channel) flows, reveals different deformation modes. Such an analysis is used to confirm the softening and stiffening effects of common treatments, such as cytochalasin D and formalin on cell deformability. In addition to an experimental analysis of leukaemia cell deformability, computational fluid dynamic simulations are used to deconvolve the role of the aforementioned flow components in the cell deformation dynamics. In general terms, the current study can be used as a guide for extracting deformation/recovery dynamics of leukaemia cell lines when exposed to various fluid dynamic conditions.

High-throughput microfluidic imaging flow cytometry.

Recently, microfluidic-based flow cytometry platforms have been shown to be powerful tools for the manipulation and analysis of single cells and micron-sized particles in flow. That said, current microfluidic flow cytometers are limited in both their analytical throughput and spatial resolution, due to their reliance on single point interrogation schemes. Conversely, high-speed imaging techniques can be applied to a wide variety of problems in which analyte molecules are manipulated at high linear velocities. Such an approach allows a detailed visualization of dynamic events through acquisition of a series of image frames captured with high temporal and spatial resolution. Herein, we describe some of the most significant recent advances in the development of multi-parametric, optofluidic imaging flow cytometry for the enumeration of complex cellular populations.

An optofluidic system with integrated microlens arrays for parallel imaging flow cytometry.

In recent years, high-speed imaging has become increasingly effective for the rapid analysis of single cells in flowing environments. Single cell imaging methods typically incorporate a minimum magnification of 10× when extracting sizing and morphological information. Although information content may be significantly enhanced by increasing magnification, this is accompanied by a corresponding reduction in field of view, and thus a decrease in the number of cells assayed per unit time. Accordingly, the acquisition of high resolution data from wide field views remains an unsolved challenge. To address this issue, we present an optofluidic flow cytometer integrating a refractive, microlens array (MLA) for imaging cells at high linear velocities, whilst maximizing the number of cells per field of view. To achieve this, we adopt an elasto-inertial approach for cell focusing within an array of parallel microfluidic channels, each equipped with a microlens. We characterize the optical performance of the microlenses in terms of image formation, magnification and resolution using both ray-tracing simulations and experimental measurements. Results demonstrate that the optofluidic platform can efficiently count and magnify micron-sized objects up to 4 times. Finally, we demonstrate the capabilities of the platform as an imaging flow cyclometer, demonstrating the efficient discrimination of hB and Jurkat cells at throughputs up to 50 000 cells per second.

Elasto-Inertial Focusing of Mammalian Cells and Bacteria Using Low Molecular, Low Viscosity PEO Solutions.

The ability to manipulate biological cells is critical in a diversity of biomedical and industrial applications. Microfluidic-based cell manipulations provide unique opportunities for sophisticated and high-throughput biological assays such as cell sorting, rare cell detection, and imaging flow cytometry. In this respect, cell focusing is an extremely useful functional operation preceding downstream biological analysis, since it allows the accurate lateral and axial positioning of cells moving through microfluidic channels, and thus enables sophisticated cell manipulations in a passive manner. Herein, we explore the utility of viscoelastic carrier fluids for enhanced elasto-inertial focusing of biological species within straight, rectangular cross section microfluidic channels. Since the investigated polymer solutions possess viscosities close to that of water and exhibit negligible shear thinning, focusing occurs over a wide range of elasticity numbers and a large range of Reynolds numbers. With a view to applications in the robust focusing of cells and bacteria, we assess and characterize the influence of accessible focusing parameters, including blockage ratio, volumetric flow rate, cell concentration, and polymer chain length.

Directed Growth of Orthorhombic Crystals in a Micropillar Array.

We report directed growth of orthorhombic crystals of potassium permanganate in spatial confinement of a micropillar array. The solution is introduced by spontaneous wicking to give a well-defined film (thickness 10-15 µm; volume ∼600 nL) and is connected to a reservoir (several microliters) that continuously "feeds" the evaporating film. When the film is supersaturated, crystals nucleate and preferentially grow in specific directions guided by one of several possible linear paths through the pillar lattice. Crystals that do not initially conform are stopped at an obstructing pillar, branch into another permitted direction, or spontaneously rotate to align with a path and continue to grow. Microspectroscopy is able to track the concentration of solute in a small region of interest (70 × 100 µm2) near to growing crystals, revealing that the solute concentration initially increases linearly beyond the solubility limit. Crystal growth near the region of interest resulted in a sharp decrease in the local solute concentration (which rapidly returns the concentration to the solubility limit), consistent with estimated diffusion time scales (<1 s for a 50 µm length scale). The ability to simultaneously track solute concentration and control crystal orientation in nanoliter samples will provide new insight into microscale dynamics of microscale crystallization.

Pillar cuvettes: capillary-filled, microliter quartz cuvettes with microscale path lengths for optical spectroscopy.

The goal of most analytical techniques is to reduce the lower limit of detection; however, it is sometimes necessary to do the opposite. High sample concentrations or samples with high molar absorptivity (e.g., dyes and metal complexes) often require multiple dilution steps or laborious sample preparation prior to spectroscopic analysis. Here, we demonstrate dilution-free, one-step UV-vis spectroscopic analysis of high concentrations of platinum(IV) hexachloride in a micropillar array, that is, "pillar cuvette". The cuvette is spontaneously filled by wicking of the liquid sample into the micropillar array. The pillar height (thus, the film thickness) defines the optical path length, which was reduced to between 10 and 20 µm in this study (3 orders of magnitude smaller than in a typical cuvette). Only one small droplet (∼2 µL) of sample is required, and the dispensed volume need not be precise or even known to the analyst for accurate spectroscopy measurements. For opaque pillars, we show that absorbance is linearly related to platinum concentration (the Beer-Lambert Law). For fully transparent or semitransparent pillars, the measured absorbance was successfully corrected for the fractional surface coverage of the pillars and the transmittance of the pillars and reference. Thus, both opaque and transparent pillars can be applied to absorbance spectroscopy of high absorptivity, microliter samples. It is also shown here that the pillar array has a useful secondary function as an integrated (in-cuvette) filter for particulates. For pillar cuvette measurements of platinum solutions spiked with 6 µm diameter polystyrene spheres, filtered and unfiltered samples gave identical spectra.