Machine Learning Based Real‐Time Image‐Guided Cell Sorting and Classification

Abstract

Cell classification based on phenotypical, spatial, and genetic information greatly advances our understanding of the physiology and pathology of biological systems. Technologies derived from next generation sequencing and fluorescent activated cell sorting are cornerstones for cell‐ and genomic‐based assays supporting cell classification and mapping. However, there exists a deficiency in technology space to rapidly isolate cells based on high content image information. Fluorescence‐activated cell sorting can only resolve cell‐to‐cell variation in fluorescence and optical scattering. Utilizing microfluidics, photonics, computation microscopy, real‐time image processing and machine learning, we demonstrate an image‐guided cell sorting and classification system possessing the high throughput of flow cytometer and high information content of microscopy. We demonstrate the utility of this technology in cell sorting based on (1) nuclear localization of glucocorticoid receptors, (2) particle binding to the cell membrane, and (3) DNA damage induced γ‐H2AX foci. © 2019 International Society for Advancement of Cytometry

Raman image-activated cell sorting

“…he advent of image-activated cell sorting [1][2][3] and imagingbased cell picking [4][5][6][7] has advanced our knowledge and exploitation of biological systems in the last decade. These foundational technologies mediate information flow between population-level analysis (flow cytometry), cell-level analysis (microscopy), and gene-level analysis (sequencing), making it possible to study, elucidate, and exploit the relations between cellular heterogeneity, phenotype, and genotype [1][2][3][4][5][6][7] .…”

“…he advent of image-activated cell sorting [1][2][3] and imagingbased cell picking [4][5][6][7] has advanced our knowledge and exploitation of biological systems in the last decade. These foundational technologies mediate information flow between population-level analysis (flow cytometry), cell-level analysis (microscopy), and gene-level analysis (sequencing), making it possible to study, elucidate, and exploit the relations between cellular heterogeneity, phenotype, and genotype [1][2][3][4][5][6][7] . Specifically, different from traditional high-content screening 8 , their ability to physically isolate target cells from large heterogeneous populations serves as a tool to identify the links between the spatial architecture of molecules within the cell (e.g., protein localization, receptor clustering, nuclear shape, cytoskeleton structure, and cell clustering) and the physiological function of the cell (e.g., proliferation, metabolism, secretion, differentiation, signaling, metastasis, and immune synapse formation) as well as for downstream characterizations (e.g., RNA sequencing and electron microscopy) and applications (e.g., cloning, directed molecular evolution, and selective breeding).…”

In Flow Cytometry, Image Is Everything

“…Zhang et al report their development of a computational method for intelligent de‐blurring of out‐of‐focus cell images in IFC by deep learning. Gu et al present their demonstration of image‐guided cell sorting together with a cell classification system and real‐time isolation of single cells based on nuclear localization of glucocorticoid receptor, particle binding to the cell membrane, and DNA damage induced γ‐H2AX foci using the instrument. Goktas et al report a combination of IFC and angle‐resolved light scattering measurements to study the morphological features of red blood cells in sickle cell patients.…”

“…They include time‐stretch imaging , frequency‐division‐multiplexed imaging , spatial–temporal transformation , temporally coded excitation , and stroboscopic fluorescence imaging , which are designed to handle sensitive image acquisition of fast‐flowing cells. The third driving factor is the availability of sophisticated digital image processing techniques, such as compressive imaging, data mining, and deep learning, powered by the rapidly evolving smartphone industry . The big data brought by IFC systems acts as a fuel for artificial neural networks that can characterize and classify cells with higher accuracy than classical image analysis algorithms.…”

Unsupervised content-preserving transformation for optical microscopy

“…The first one is image classification 6,7 where deep neural networks are trained to classify an image in terms of its visual content. Such applications include cell sorting 8,9 , morphology recognition 10,11 , state identification 12 and pathological diagnosis 13,14 . The other one is image transformation whose outputs are still images but with highlighted or previously inaccessible information, aiming to visualize imperceptible structures and latent patterns, as well as expand the design space of common imaging systems 15 .…”