Cell segmentation methods for label-free contrast microscopy: review and comprehensive comparison

Abstract

Background Because of its non-destructive nature, label-free imaging is an important strategy for studying biological processes. However, routine microscopic techniques like phase contrast or DIC suffer from shadow-cast artifacts making automatic segmentation challenging. The aim of this study was to compare the segmentation efficacy of published steps of segmentation work-flow (image reconstruction, foreground segmentation, cell detection (seed-point extraction) and cell (instance) segmentation) on a dataset of the same cells from multiple contrast microscopic modalities. Results We built a collection of routines aimed at image segmentation of viable adherent cells grown on the culture dish acquired by phase contrast, differential interference contrast, Hoffman modulation contrast and quantitative phase imaging, and we performed a comprehensive comparison of available segmentation methods applicable for label-free data. We demonstrated that it is crucial to perform the image reconstruction step, enabling the use of segmentation methods originally not applicable on label-free images. Further we compared foreground segmentation methods (thresholding, feature-extraction, level-set, graph-cut, learning-based), seed-point extraction methods (Laplacian of Gaussians, radial symmetry and distance transform, iterative radial voting, maximally stable extremal region and learning-based) and single cell segmentation methods. We validated suitable set of methods for each microscopy modality and published them online. Conclusions We demonstrate that image reconstruction step allows the use of segmentation methods not originally intended for label-free imaging. In addition to the comprehensive comparison of methods, raw and reconstructed annotated data and Matlab codes are provided. Electronic supplementary material The online version of this article (10.1186/s12859-019-2880-8) contains supplementary material, which is available to authorized users.

A Self-Supervised Machine Learning Approach for Objective Live Cell Segmentation and Analysis

“…Furthermore, there has been an increased appreciation for the morphological information label-free approaches can provide as a result of algorithmic-based phenotyping. 20-22…”

Pinpointing Cell Identity in Time and Space

“…In addition to spot detection, spatially resolved transcriptomics necessitates the ability to distinguish individual and adjacent cells from each other, and a way to characterize the distribution of FISH spots across individual cells. While there are a large number of cell segmentation tools available (reviewed in Meijering, 2012 and Vicar et al, 2019 ), the automated segmentation of densely packed cells and nuclei, either in a cultured monolayer or intact tissue sections remains a challenge. Potential solutions to this may also lie in machine learning and network-based approaches (for example Al-Kofahi et al, 2018 ; Schmidt et al, 2018 ; Berg et al, 2019 ).…”

Bioimage Analysis and Cell Motility

“…Label-free imaging such as brightfield or phase-contrast microscopy is even less invasive but introduces halo-like artifacts that hamper segmentation; in fact, no feature by itself seems characteristic enough of a cell observed under transmitted light. 91 Challenges like this have popularized interactive machine learning (IML) algorithms ( Figure 2 ; Table 1 ), which use non-linear classifiers (e.g., random forest) to select the best combination of features from a pre-set collection of filters (see above: e.g., edge, texture) by training on user-annotated data. 92 , 93 On the other hand, rather than working with a predetermined set of features like IML, deep learning and artificial neural networks (ANNs) tailor their own filters to the training set from a rather general template that includes a range of non-linear mappings; 94 that is, they look for good features automatically, but, in exchange, the underlying assumptions are practically inaccessible.…”