Today, Convolutional Neural Networks (CNNs) are the leading method for image denoising. They are traditionally trained on pairs of images, which are often hard to obtain for practical applications. This motivates self-supervised training methods such as Noise2Void (N2V) that operate on single noisy images. Self-supervised methods are, unfortunately, not competitive with models trained on image pairs. Here, we present Probabilistic Noise2Void (PN2V), a method to train CNNs to predict per-pixel intensity distributions. Combining these with a suitable description of the noise, we obtain a complete probabilistic model for the noisy observations and true signal in every pixel. We evaluate PN2V on publicly available microscopy datasets, under a broad range of noise regimes, and achieve competitive results with respect to supervised state-of-the-art methods.
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.
Cell viability and cytotoxicity assays are highly important for drug screening and cytotoxicity tests of antineoplastic or other therapeutic drugs. Even though biochemical-based tests are very helpful to obtain preliminary preview, their results should be confirmed by methods based on direct cell death assessment. In this study, time-dependent changes in quantitative phase-based parameters during cell death were determined and methodology useable for rapid and label-free assessment of direct cell death was introduced. The goal of our study was distinction between apoptosis and primary lytic cell death based on morphologic features. We have distinguished the lytic and non-lytic type of cell death according to their end-point features (Dance of Death typical for apoptosis versus swelling and membrane rupture typical for all kinds of necrosis common for necroptosis, pyroptosis, ferroptosis and accidental cell death). Our method utilizes Quantitative Phase Imaging (QPI) which enables the time-lapse observation of subtle changes in cell mass distribution. According to our results, morphological and dynamical features extracted from QPI micrographs are suitable for cell death detection (76% accuracy in comparison with manual annotation). Furthermore, based on QPI data alone and machine learning, we were able to classify typical dynamical changes of cell morphology during both caspase 3,7-dependent and -independent cell death subroutines. The main parameters used for label-free detection of these cell death modalities were cell density (pg/pixel) and average intensity change of cell pixels further designated as Cell Dynamic Score (CDS). To the best of our knowledge, this is the first study introducing CDS and cell density as a parameter typical for individual cell death subroutines with prediction accuracy 75.4% for caspase 3,7-dependent and -independent cell death.
We focused on the biomechanical and morphological characteristics of prostate cancer cells and their changes resulting from the effect of docetaxel, cisplatin, and long-term zinc supplementation. Cell population surviving the treatment was characterized as follows: cell stiffness was assessed by atomic force microscopy, cell motility and invasion capacity were determined by colony forming assay, wound healing assay, coherence-controlled holographic microscopy, and real-time cell analysis. Cells of metastatic origin exhibited lower height than cells derived from the primary tumour. Cell dry mass and CAV1 gene expression followed similar trends as cell stiffness. Docetaxel- and cisplatin-surviving cells had higher stiffness, and decreased motility and invasive potential as compared to non-treated cells. This effect was not observed in zinc(II)-treated cells. We presume that cell stiffness changes may represent an important overlooked effect of cisplatin-based anti-cancer drugs. Atomic force microscopy and confocal microscopy data images used in our study are available for download in the Zenodo repository (https://zenodo.org/, Digital Object Identifiers:10.5281/zenodo.1494935).
From the very beginnings of radiotherapy, a crucial question persists with how to target the radiation effectiveness into the tumor while preserving surrounding tissues as undamaged as possible. One promising approach is to selectively pre-sensitize tumor cells by metallic nanoparticles. However, though the “physics” behind nanoparticle-mediated radio-interaction has been well elaborated, practical applications in medicine remain challenging and often disappointing because of limited knowledge on biological mechanisms leading to cell damage enhancement and eventually cell death. In the present study, we analyzed the influence of different nanoparticle materials (platinum (Pt), and gold (Au)), cancer cell types (HeLa, U87, and SKBr3), and doses (up to 4 Gy) of low-Linear Energy Transfer (LET) ionizing radiation (γ- and X-rays) on the extent, complexity and reparability of radiation-induced γH2AX + 53BP1 foci, the markers of double stand breaks (DSBs). Firstly, we sensitively compared the focus presence in nuclei during a long period of time post-irradiation (24 h) in spatially (three-dimensionally, 3D) fixed cells incubated and non-incubated with Pt nanoparticles by means of high-resolution immunofluorescence confocal microscopy. The data were compared with our preliminary results obtained for Au nanoparticles and recently published results for gadolinium (Gd) nanoparticles of approximately the same size (2–3 nm). Next, we introduced a novel super-resolution approach—single molecule localization microscopy (SMLM)—to study the internal structure of the repair foci. In these experiments, 10 nm Au nanoparticles were used that could be also visualized by SMLM. Altogether, the data show that different nanoparticles may or may not enhance radiation damage to DNA, so multi-parameter effects have to be considered to better interpret the radiosensitization. Based on these findings, we discussed on conclusions and contradictions related to the effectiveness and presumptive mechanisms of the cell radiosensitization by nanoparticles. We also demonstrate that SMLM offers new perspectives to study internal structures of repair foci with the goal to better evaluate potential differences in DNA damage patterns.
Black phosphorus (BP) belongs to a group of 2D nanomaterials and nowadays attracts constantly increasing attention. Parallel to the growing utilization of BP nanomaterial increase also the requirements for the thorough comprehension of its potential impact on human and animal health. The aim of this study was to compare and discuss five assays commonly used for the cytotoxicity assessments of nanomaterials with a special focus on BP nanoparticles. A comprehensive survey of factors and pitfalls is provided that should be accounted for when assessing their toxicity and pointed to their inconsistency. BP might introduce various levels of interference during toxicity assessments depending on its concentration applied. More importantly, the BP toxicity evaluation was found to be influenced by the nature of assay chosen. These are based on different principles and do not have to assess all the cellular events equally. A commercial assay based on the measurement of protease activity was identified to be the most suitable for the BP toxicity assessment. Further, the benefit of time‐lapse quantitative phase imaging for nanomaterial toxicity evaluation was highlighted. Unlike the conventional assessments it provides real‐time analysis of the processes accompanying BP administration and enables to understand them deeper and in the context.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.