Information-Distilled Generative Label-Free Morphological Profiling Encodes Cellular Heterogeneity.

Image-based cytometry faces challenges due to technical variations arising from different experimental batches and conditions, such as differences in instrument configurations or image acquisition protocols, impeding genuine biological interpretation of cell morphology. Existing solutions, often necessitating extensive pre-existing data knowledge or control samples across batches, have proved limited, especially with complex cell image data. To overcome this, "Cyto-Morphology Adversarial Distillation" (CytoMAD), a self-supervised multi-task learning strategy that distills biologically relevant cellular morphological information from batch variations, is introduced to enable integrated analysis across multiple data batches without complex data assumptions or extensive manual annotation. Unique to CytoMAD is its "morphology distillation", symbiotically paired with deep-learning image-contrast translation-offering additional interpretable insights into label-free cell morphology. The versatile efficacy of CytoMAD is demonstrated in augmenting the power of biophysical imaging cytometry. It allows integrated label-free classification of human lung cancer cell types and accurately recapitulates their progressive drug responses, even when trained without the drug concentration information. CytoMAD  also allows joint analysis of tumor biophysical cellular heterogeneity, linked to epithelial-mesenchymal plasticity, that standard fluorescence markers overlook. CytoMAD can substantiate the wide adoption of biophysical cytometry for cost-effective diagnosis and screening.

Morphological profiling by high-throughput single-cell biophysical fractometry.

Complex and irregular cell architecture is known to statistically exhibit fractal geometry, i.e., a pattern resembles a smaller part of itself. Although fractal variations in cells are proven to be closely associated with the disease-related phenotypes that are otherwise obscured in the standard cell-based assays, fractal analysis with single-cell precision remains largely unexplored. To close this gap, here we develop an image-based approach that quantifies a multitude of single-cell biophysical fractal-related properties at subcellular resolution. Taking together with its high-throughput single-cell imaging performance (~10,000 cells/sec), this technique, termed single-cell biophysical fractometry, offers sufficient statistical power for delineating the cellular heterogeneity, in the context of lung-cancer cell subtype classification, drug response assays and cell-cycle progression tracking. Further correlative fractal analysis shows that single-cell biophysical fractometry can enrich the standard morphological profiling depth and spearhead systematic fractal analysis of how cell morphology encodes cellular health and pathological conditions.

StarmapVis: An interactive and narrative visualisation tool for single-cell and spatial data.

Current single-cell visualisation techniques project high dimensional data into 'map' views to identify high-level structures such as cell clusters and trajectories. New tools are needed to allow the transversal through the high dimensionality of single-cell data to explore the single-cell local neighbourhood. StarmapVis is a convenient web application displaying an interactive downstream analysis of single-cell expression or spatial transcriptomic data. The concise user interface is powered by modern web browsers to explore the variety of viewing angles unavailable to 2D media. Interactive scatter plots display clustering information, while the trajectory and cross-comparison among different coordinates are displayed in connectivity networks. Automated animation of camera view is a unique feature of our tool. StarmapVis also offers a useful animated transition between two-dimensional spatial omic data to three-dimensional single cell coordinates. The usability of StarmapVis is demonstrated by four data sets, showcasing its practical usability. StarmapVis is available at: https://holab-hku.github.io/starmapVis.

Optofluidic imaging meets deep learning: from merging to emerging.

Propelled by the striking advances in optical microscopy and deep learning (DL), the role of imaging in lab-on-a-chip has dramatically been transformed from a silo inspection tool to a quantitative "smart" engine. A suite of advanced optical microscopes now enables imaging over a range of spatial scales (from molecules to organisms) and temporal window (from microseconds to hours). On the other hand, the staggering diversity of DL algorithms has revolutionized image processing and analysis at the scale and complexity that were once inconceivable. Recognizing these exciting but overwhelming developments, we provide a timely review of their latest trends in the context of lab-on-a-chip imaging, or coined optofluidic imaging. More importantly, here we discuss the strengths and caveats of how to adopt, reinvent, and integrate these imaging techniques and DL algorithms in order to tailor different lab-on-a-chip applications. In particular, we highlight three areas where the latest advances in lab-on-a-chip imaging and DL can form unique synergisms: image formation, image analytics and intelligent image-guided autonomous lab-on-a-chip. Despite the on-going challenges, we anticipate that they will represent the next frontiers in lab-on-a-chip imaging that will spearhead new capabilities in advancing analytical chemistry research, accelerating biological discovery, and empowering new intelligent clinical applications.

Deep-learning-assisted biophysical imaging cytometry at massive throughput delineates cell population heterogeneity.

The association of the intrinsic optical and biophysical properties of cells to homeostasis and pathogenesis has long been acknowledged. Defining these label-free cellular features obviates the need for costly and time-consuming labelling protocols that perturb the living cells. However, wide-ranging applicability of such label-free cell-based assays requires sufficient throughput, statistical power and sensitivity that are unattainable with current technologies. To close this gap, we present a large-scale, integrative imaging flow cytometry platform and strategy that allows hierarchical analysis of intrinsic morphological descriptors of single-cell optical and mass density within a population of millions of cells. The optofluidic cytometry system also enables the synchronous single-cell acquisition of and correlation with fluorescently labeled biochemical markers. Combined with deep neural network and transfer learning, this massive single-cell profiling strategy demonstrates the label-free power to delineate the biophysical signatures of the cancer subtypes, to detect rare populations of cells in the heterogeneous samples (10-5), and to assess the efficacy of targeted therapeutics. This technique could spearhead the development of optofluidic imaging cell-based assays that stratify the underlying physiological and pathological processes based on the information-rich biophysical cellular phenotypes.

PARC: ultrafast and accurate clustering of phenotypic data of millions of single cells.

MOTIVATION: New single-cell technologies continue to fuel the explosive growth in the scale of heterogeneous single-cell data. However, existing computational methods are inadequately scalable to large datasets and therefore cannot uncover the complex cellular heterogeneity. RESULTS: We introduce a highly scalable graph-based clustering algorithm PARC-Phenotyping by Accelerated Refined Community-partitioning-for large-scale, high-dimensional single-cell data (>1 million cells). Using large single-cell flow and mass cytometry, RNA-seq and imaging-based biophysical data, we demonstrate that PARC consistently outperforms state-of-the-art clustering algorithms without subsampling of cells, including Phenograph, FlowSOM and Flock, in terms of both speed and ability to robustly detect rare cell populations. For example, PARC can cluster a single-cell dataset of 1.1 million cells within 13 min, compared with >2 h for the next fastest graph-clustering algorithm. Our work presents a scalable algorithm to cope with increasingly large-scale single-cell analysis. AVAILABILITY AND IMPLEMENTATION: https://github.com/ShobiStassen/PARC. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Parallelized volumetric fluorescence microscopy with a reconfigurable coded incoherent light-sheet array.

Parallelized fluorescence imaging has been a long-standing pursuit that can address the unmet need for a comprehensive three-dimensional (3D) visualization of dynamical biological processes with minimal photodamage. However, the available approaches are limited to incomplete parallelization in only two dimensions or sparse sampling in three dimensions. We hereby develop a novel fluorescence imaging approach, called coded light-sheet array microscopy (CLAM), which allows complete parallelized 3D imaging without mechanical scanning. Harnessing the concept of an "infinity mirror", CLAM generates a light-sheet array with controllable sheet density and degree of coherence. Thus, CLAM circumvents the common complications of multiple coherent light-sheet generation in terms of dedicated wavefront engineering and mechanical dithering/scanning. Moreover, the encoding of multiplexed optical sections in CLAM allows the synchronous capture of all sectioned images within the imaged volume. We demonstrate the utility of CLAM in different imaging scenarios, including a light-scattering medium, an optically cleared tissue, and microparticles in fluidic flow. CLAM can maximize the signal-to-noise ratio and the spatial duty cycle, and also provides a further reduction in photobleaching compared to the major scanning-based 3D imaging systems. The flexible implementation of CLAM regarding both hardware and software ensures compatibility with any light-sheet imaging modality and could thus be instrumental in a multitude of areas in biological research.