Abstract:Automated cell classification in cancer biology is a challenging topic in computer vision and machine learning research. Breast cancer is the most common malignancy in women that usually involves phenotypically diverse populations of breast cancer cells and an heterogeneous stroma. In recent years, automated microscopy technologies are allowing the study of live cells over extended periods of time, simplifying the task of compiling large image databases. For instance, there have been several studies oriented towards building machine learning systems capable of automatically classifying images of different cell types (i.e. motor neurons, stem cells). In this work we were interested in classifying breast cancer cells as live or dead, based on a set of automatically retrieved morphological characteristics using image processing techniques. Our hypothesis is that live-dead classification can be performed without any staining and using only bright-field images as input. We tackled this problem using the JIMT-1 breast cancer cell line that grows as an adherent monolayer. First, a vast image set composed by JIMT-1 human breast cancer cells that had been exposed to a chemotherapeutic drug treatment (doxorubicin and paclitaxel) or vehicle control was compiled. Next, several classifiers were trained based on well-known convolutional neural networks (CNN) backbones to perform supervised classification using labels obtained from fluorescence microscopy images associated with each bright-field image. Model performances were evaluated and compared on a large number of bright-field images. The best model reached an AUC = 0.941 for classifying breast cancer cells without treatment. Furthermore, it reached AUC = 0.978 when classifying breast cancer cells under drug treatment. Our results highlight the potential of machine learning and computational image analysis to build new diagnosis tools that benefit the biomedical field by reducing cost, time, and stimulating work reproducibility. More importantly, we analyzed the way our classifiers clusterize bright-field images in the learned high-dimensional embedding and linked these groups to salient visual characteristics in live-dead cell biology observed by trained experts.

In Vivo Identification of Apoptotic and Extracellular Vesicle-Bound Live Cells Using Image-Based Deep Learning

Consensus guidelines for the detection of immunogenic cell death.

Single Extracellular Vesicle Analysis Using Droplet Microfluidics.

Transformer-based spatial–temporal detection of apoptotic cell death in live-cell imaging

Rapid Enrichment and Sensitive Detection of Extracellular Vesicles Through Measuring the Phospholipids and Transmembrane Protein in a Microfluidic Chip.

Assessing Extracellular Vesicle Turnover In vivo Using Highly Sensitive Phosphatidylserine-Binding Reagents

Detection and Quantification of Extracellular Vesicles via FACS: Membrane Labeling Matters!

Live-dead assay on unlabeled cells using phase imaging with computational specificity

Real-time Detection of Circulating Apoptotic Cells by in Vivo Flow Cytometry

Intelligent Probabilistic System for Digital Tracing Cellular Origin of Individual Clinical Extracellular Vesicles

Evaluation of a Deep Learning Based Approach to Computational Label Free Cell Viability Quantification

Unsupervised Machine Learning-Based Clustering of Nanosized Fluorescent Extracellular Vesicles

Stain-free approach to determine and monitor cell heath using supervised and unsupervised image-based deep learning

Potential-Resolved Electrochemiluminescence Nanoprobes for Visual Apoptosis Evaluation at Single-Cell Level

Apoptosis detection for adherent cell populations in time-lapse phase-contrast microscopy images.

Detection of the interactions of tumour derived extracellular vesicles with immune cells is dependent on EV‐labelling methods

A Real‐Time Image‐Based Approach to Distinguish and Discriminate Apoptosis from Necrosis

Large Extracellular Vesicles Can be Characterised by Multiplex Labelling Using Imaging Flow Cytometry

AI-based Apoptosis Cell Classification Using Phase-contrast Images of K562 Cells

Learning deep features for dead and living breast cancer cell classification without staining

Rapid Assessment of Biomarkers on Single Extracellular Vesicles Using ‘Catch and Display’ on Ultrathin Nanoporous Silicon Nitride Membranes