ACDC: Automated Cell Detection and Counting for Time-lapse Fluorescence Microscopy

AbstractAdvances in microscopy imaging technologies have enabled the visualization of live-cell dynamic processes using time-lapse microscopy imaging. However, modern methods exhibit several limitations related to the training phases and to time constraints, hindering their application in the laboratory practice. In this work, we present a novel method, named Automated Cell Detection and Counting (ACDC), designed for activity detection of fluorescent labeled cell nuclei in time-lapse microscopy. ACDC overcomes the limitations of the literature methods, by first applying bilateral filtering on the original image to smooth the input cell images while preserving edge sharpness, and then by exploiting the watershed transform and morphological filtering. Moreover, ACDC represents a feasible solution for the laboratory practice, as it can leverage multi-core architectures in computer clusters to efficiently handle large-scale imaging datasets. Indeed, our Parent-Workers implementation of ACDC allows to obtain up to a 3.7× speed-up compared to the sequential counterpart. ACDC was tested on two distinct cell imaging datasets to assess its accuracy and effectiveness on images with different characteristics. We achieved an accurate cell-count and nuclei segmentation without relying on large-scale annotated datasets, a result confirmed by the average Dice Similarity Coefficients of 76.84 and 88.64 and the Pearson coefficients of 0.99 and 0.96, calculated against the manual cell counting, on the two tested datasets.

Download Full-text

ACDC: Automated Cell Detection and Counting for Time-Lapse Fluorescence Microscopy

Applied Sciences ◽

10.3390/app10186187 ◽

2020 ◽

Vol 10 (18) ◽

pp. 6187

Author(s):

Leonardo Rundo ◽

Andrea Tangherloni ◽

Darren R. Tyson ◽

Riccardo Betta ◽

Carmelo Militello ◽

...

Keyword(s):

Large Scale ◽

Time Lapse ◽

Cell Counting ◽

Cell Detection ◽

Cell Nuclei ◽

Similarity Coefficients ◽

Laboratory Practice ◽

Microscopy Imaging ◽

Time Lapse Microscopy ◽

Detection And Counting

Advances in microscopy imaging technologies have enabled the visualization of live-cell dynamic processes using time-lapse microscopy imaging. However, modern methods exhibit several limitations related to the training phases and to time constraints, hindering their application in the laboratory practice. In this work, we present a novel method, named Automated Cell Detection and Counting (ACDC), designed for activity detection of fluorescent labeled cell nuclei in time-lapse microscopy. ACDC overcomes the limitations of the literature methods, by first applying bilateral filtering on the original image to smooth the input cell images while preserving edge sharpness, and then by exploiting the watershed transform and morphological filtering. Moreover, ACDC represents a feasible solution for the laboratory practice, as it can leverage multi-core architectures in computer clusters to efficiently handle large-scale imaging datasets. Indeed, our Parent-Workers implementation of ACDC allows to obtain up to a 3.7× speed-up compared to the sequential counterpart. ACDC was tested on two distinct cell imaging datasets to assess its accuracy and effectiveness on images with different characteristics. We achieved an accurate cell-count and nuclei segmentation without relying on large-scale annotated datasets, a result confirmed by the average Dice Similarity Coefficients of 76.84 and 88.64 and the Pearson coefficients of 0.99 and 0.96, calculated against the manual cell counting, on the two tested datasets.

Download Full-text

Automated Segmentation, Classification, and Tracking of Cancer Cell Nuclei in Time-Lapse Microscopy

IEEE Transactions on Biomedical Engineering ◽

10.1109/tbme.2006.870201 ◽

2006 ◽

Vol 53 (4) ◽

pp. 762-766 ◽

Cited By ~ 228

Author(s):

X. Chen ◽

X. Zhou ◽

S.T.C. Wong

Keyword(s):

Cancer Cell ◽

Time Lapse ◽

Automated Segmentation ◽

Cell Nuclei ◽

Time Lapse Microscopy

Download Full-text

Methodological Implications on Quantitative Studies of Cytocompatibility in Direct Contact with Bioceramic Surfaces

Key Engineering Materials ◽

10.4028/www.scientific.net/kem.493-494.325 ◽

2011 ◽

Vol 493-494 ◽

pp. 325-330 ◽

Cited By ~ 1

Author(s):

J.A. Cortês ◽

Elena Mavropoulos ◽

Moema Hausen ◽

Alexandre Rossi ◽

J.M. Granjeiro ◽

...

Keyword(s):

Cell Adhesion ◽

Crystal Violet ◽

Fluorescent Dyes ◽

Ceramic Materials ◽

Neutral Red ◽

Cell Counting ◽

Cell Detection ◽

Porous Hydroxyapatite ◽

Detection And Counting ◽

Direct Cell

Cell adhesion, proliferation and differentiation are important specific parameters to be evaluated on biocompatibility studies of candidate biomaterials for clinical applications. Several different methodologies have been employed to study, both qualitative and quantitatively, the direct interactions of ceramic materials with cultured mammal and human cells. However, while quantitatively evaluating cell density, viability and metabolic responses to test materials, several methodological challenges may arise, either by impairing the use of some widely applied techniques, or by generating false or conflicting results. In this work, we tested the inherent interference of different representative calcium phosphate ceramic surfaces (stoichiometric dense and porous hydroxyapatite (HA) and cation-substituted apatite tablets) on different tests for quantitative evaluation of osteoblast adhesion and metabolism, either based on direct cell counting after trypsinization, colorimetric assays (XTT, Neutral Red and Crystal Violet) and fluorescence microscopy. Cell adhesion estimation after trypsinization was highly dependent on the time of treatment, and the group with the highest level of estimated adhesion was inverted from 5 to 20 minutes of exposition to trypsin. Both dense and porous HA samples presented high levels of background adsorption of the Crystal Violet dye, impairing cell detection. HA surfaces also were able to adsorb high levels of fluorescent dyes (DAPI and phalloidin-TRITC), generating backgrounds which, in the case of porous HA, impaired cell detection and counting by image processing software (Image Pro Plus 6.0). We conclude that the choice for the most suitable method for cell detection and estimation is highly dependent on very specific characteristics of the studied material, and methodological adaptations on well established protocols must always be carefully taken on consideration.

Download Full-text

Enhancing Multi-tissue and Multi-scale Cell Nuclei Segmentation with Deep Metric Learning

Applied Sciences ◽

10.3390/app10020615 ◽

2020 ◽

Vol 10 (2) ◽

pp. 615 ◽

Cited By ~ 2

Author(s):

Tomas Iesmantas ◽

Agne Paulauskaite-Taraseviciene ◽

Kristina Sutiene

Keyword(s):

Deep Learning ◽

Large Scale ◽

Metric Learning ◽

Cell Nuclei ◽

Similarity Coefficients ◽

Clinical Practices ◽

Nuclei Segmentation ◽

Wide Range ◽

Triplet Loss ◽

Deep Metric Learning

(1) Background: The segmentation of cell nuclei is an essential task in a wide range of biomedical studies and clinical practices. The full automation of this process remains a challenge due to intra- and internuclear variations across a wide range of tissue morphologies, differences in staining protocols and imaging procedures. (2) Methods: A deep learning model with metric embeddings such as contrastive loss and triplet loss with semi-hard negative mining is proposed in order to accurately segment cell nuclei in a diverse set of microscopy images. The effectiveness of the proposed model was tested on a large-scale multi-tissue collection of microscopy image sets. (3) Results: The use of deep metric learning increased the overall segmentation prediction by 3.12% in the average value of Dice similarity coefficients as compared to no metric learning. In particular, the largest gain was observed for segmenting cell nuclei in H&E -stained images when deep learning network and triplet loss with semi-hard negative mining were considered for the task. (4) Conclusion: We conclude that deep metric learning gives an additional boost to the overall learning process and consequently improves the segmentation performance. Notably, the improvement ranges approximately between 0.13% and 22.31% for different types of images in the terms of Dice coefficients when compared to no metric deep learning.

Download Full-text

Microscopy quantification of microbial birth and death dynamics

10.1101/324269 ◽

2018 ◽

Cited By ~ 3

Author(s):

Samuel F. M. Hart ◽

David Skelding ◽

Adam J. Waite ◽

Justin Burton ◽

Li Xie ◽

...

Keyword(s):

Fluorescent Proteins ◽

Time Lapse ◽

Microbial Evolution ◽

Cell Counting ◽

Live Cells ◽

Nutrient Concentrations ◽

Death Rates ◽

Time Lapse Microscopy ◽

Wide Range ◽

Total Fluorescence

AbstractMicrobes live in dynamic environments where nutrient concentrations fluctuate. Quantifying fitness (birth and death) in a wide range of environments is critical for understanding microbial evolution as well as ecological interactions where one species alters the fitness of another. Here, using high-throughput time-lapse microscopy, we have quantified howSaccharomyces cerevisiaemutants incapable of synthesizing an essential metabolite grow or die in various concentrations of the required metabolite. We establish that cells normally expressing fluorescent proteins lose fluorescence upon death and that the total fluorescence in an imaging frame is proportional to the number of live cells even when cells form multiple layers. We validate our microscopy approach of measuring birth and death rates using flow cytometry, cell counting, and chemostat culturing. For lysine-requiring cells, very low concentrations of lysine are not detectably consumed and do not support cell birth, but delay the onset of death phase and reduce the death rate. In contrast, in low hypoxanthine, hypoxanthine-requiring cells can produce new cells, yet also die faster than in the absence of hypoxanthine. For both strains, birth rates under various metabolite concentrations are better described by the sigmoidal-shaped Moser model than the well-known Monod model, while death rates depend on the metabolite concentration and can vary with time. Our work reveals how time-lapse microscopy can be used to discover non-intuitive microbial dynamics and to quantify growth rates in many environments.

Download Full-text

BCData: A Large-Scale Dataset and Benchmark for Cell Detection and Counting

Medical Image Computing and Computer Assisted Intervention – MICCAI 2020 - Lecture Notes in Computer Science ◽

10.1007/978-3-030-59722-1_28 ◽

2020 ◽

pp. 289-298

Author(s):

Zhongyi Huang ◽

Yao Ding ◽

Guoli Song ◽

Lin Wang ◽

Ruizhe Geng ◽

...

Keyword(s):

Large Scale ◽

Cell Detection ◽

Large Scale Dataset ◽

Detection And Counting

Download Full-text

An Advanced Framework for Time-lapse Microscopy Image Analysis

10.1101/2020.09.21.303800 ◽

2020 ◽

Author(s):

Qibing Jiang ◽

Praneeth Sudalagunta ◽

Mark B. Meads ◽

Khandakar Tanvir Ahmed ◽

Tara Rutkowski ◽

...

Keyword(s):

Image Analysis ◽

Drug Response ◽

Time Lapse ◽

Response To Therapy ◽

Patient Specific ◽

Microscopy Imaging ◽

Time Lapse Microscopy ◽

Link Type ◽

Microscopy Image ◽

Microscopy Image Analysis

ABSTRACTTime-lapse microscopy is a powerful technique that generates large volumes of image-based information to quantify the behaviors of cell populations. This method has been applied to cancer studies to estimate the drug response for precision medicine and has great potential to address inter-patient (or intertumoral) heterogeneity. A couple of algorithms exist to analyze time-lapse microscopy images; however, most deal with very high-resolution images involving few cells (typically cell lines). There are currently no advanced and efficient computational frameworks available to process large-scale time-lapse microscopy imaging data to estimate patient-specific response to therapy based on a large population of primary cells. In this paper, we propose a robust and user-friendly pipeline to preprocess the images and track the behaviors of thousands of cancer cells simultaneously for a better drug response prediction of cancer patients.Availability and ImplementationSource code is available at: https://github.com/CompbioLabUCF/CellTrackACM Reference FormatQibing Jiang, Praneeth Sudalagunta, Mark B. Meads, Khandakar Tanvir Ahmed, Tara Rutkowski, Ken Shain, Ariosto S. Silva, and Wei Zhang. 2020. An Advanced Framework for Time-lapse Microscopy Image Analysis. In Proceedings of BioKDD: 19th International Workshop on Data Mining In Bioinformatics (BioKDD). ACM, New York, NY, USA, 8 pages. https://doi.org/10.1145/nnnnnnn.nnnnnnn

Download Full-text

A System for Spatio-Temporal Cell Detection and Segmentation in Time-Lapse Microscopy

10.1109/bibm52615.2021.9669421 ◽

2021 ◽

Author(s):

Sokratis Makrogiannis ◽

Nagasoujanya Annasamudram ◽

Yibing Wang ◽

Hector Miranda ◽

Keni Zheng

Keyword(s):

Time Lapse ◽

Cell Detection ◽

Time Lapse Microscopy ◽

Spatio Temporal

Download Full-text

Image and video analysis techniques for cellular microscopy

10.32469/10355/46443 ◽

2014 ◽

Author(s):

◽

Ilker Ersoy

Keyword(s):

Level Set ◽

Image Data ◽

Time Lapse ◽

Image Features ◽

Biological Knowledge ◽

Cell Detection ◽

Automated Classification ◽

Stepping Stones ◽

Cell Dynamics ◽

Microscopy Imaging

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] Advances in automated digital microscopy imaging made it possible to produce multi-dimensional image data that can capture dynamic characteristics of sub-cellular and cellular structures. Biologists routinely produce large volumes of bioimage time lapse data that necessitates automated algorithms for unbiased and repeatable quantitative analysis. These algorithms are the stepping stones in bioimage informatics to turn the image data into biological knowledge. Unique challenges posed by different imaging modalities and cell dynamics require a combination of accurate detection, segmentation, classification and tracking approaches tailored to address and exploit particular image characteristics. In this dissertation, we present algorithms for the analysis of microscopy image sequences to address these challenges. We propose a level set active contour approach to address accurate segmentation in phase-contrast as well as brightfield microscopy imaging that utilizes edge profiles. Our approach significantly outperforms traditional level set approaches. We show the applications of our approach to cell spreading analysis and red blood cell analysis with robust solutions for cell detection to delineate clustered cells. We also present two studies for automated classification of cells in fluorescence microscopy emphasizing the importance of choosing image features for the specific problem. Lastly, we present a fully automated cell detection and tracking approach tailored for muscle satellite cells that enables efficient and unbiased analysis of factors that promote cell motility.

Download Full-text