Image segmentation of mesenchymal stem cells in diverse culturing conditions

Afridi, Muhammad Jamal; Liu, Chun; Chan, Christina; Baek, Seungik; Liu, Xiaoming

doi:10.1109/wacv.2014.6836058

Cited by 3 publications

(5 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hence, global thresholding cannot be effective for segmentation purpose, making it essential to extract local areas to reduce the complexity of thresholding. To do so, variance in intensity of cell in the image and variance in intensity of background pixels can be measured as cell-pixel variance V c and backgroundpixel variance V b respectively, to quantify the complexity which arises from intensity variation 31 . The high value of either V c or V b or both shows the overlapping in the cell and background intensities, thereby increasing the difficulty or complexity in the segmentation process.…”

Section: Global Versus Local Thresholdingmentioning

confidence: 99%

“…For defining the threshold, these approaches have exploited different features of the image as mentioned. Out of all these various approaches, it has been found that the approaches giving results close to the true value are Huang, Li and BEAS, with the BEAS being the most effective than other two 31 . Furthermore, BEAS method is reportedly having more advantages and reliability than many automated methods such as Fast Fourier Transform-Radial Sum 38 (FFTRS) and gradient-based approaches 39 .…”

Section: Thresholdingmentioning

confidence: 99%

“…Hence, the shifted part will define the accuracy of the threshold value and it is dependent on the mean and standard deviation only. However, for certain images, the cell distribution may have same mean and standard deviation, but have different optimal threshold values 31 . Afridi et al 31 proposed an improvement to this method by exploiting the rate of increase of pixel intensity from low values to peak values as another characterizing factor, because for same mean and standard deviation, this rate  can be different 31 .…”

Section: Thresholdingmentioning

confidence: 99%

“…However, for certain images, the cell distribution may have same mean and standard deviation, but have different optimal threshold values 31 . Afridi et al 31 proposed an improvement to this method by exploiting the rate of increase of pixel intensity from low values to peak values as another characterizing factor, because for same mean and standard deviation, this rate  can be different 31 . The threshold in this method is calculated for different patches of images through pixel intensities x and corresponding pixel count y, considering only the distribution towards the left of the most frequent intensity (m f ).…”

Section: Thresholdingmentioning

confidence: 99%

See 3 more Smart Citations

Extracting Cell and Fibre Orientation from Microscopic Images:Computational Methods and Tools

Sharma

2016

Current Science

View full text Add to dashboard Cite

Orientation of cell and its actin fibres dictate various properties of the cell and its associated tissue including morphology, cell migration and response of cell to external stimuli. While the vital role of cell and actin fibre orientation has been demonstrated in various studies and across many cell-types, extracting this information from microscopy images is a challenging task. Although several methods have been developed to quantify alignment of a cell and its actin fibres from experimental images, all of them differ at one step or the other. We highlight the importance of orientation of cell and associated fibres and present a generic scheme of various approaches to assimilate the widely used methods. We have presented a systematic approach to determine cell and fibre orientation. The study will improve our understanding of the core processes and will also help in development of high throughput imaging methods to extract information from experimental images.Keywords: BEAS, cell orientation, DFT, fibre orientation, FFT, image processing, thresholding.IN the study of biological phenomena, cell is at a very fundamental level, as it is the smallest building block of all organisms. In fact, it is the smallest form of life that exhibits all the basic traits that form life such as birth, ingestion, digestion, protection and finally death. Cell during its life cycle experiences a number of mechanical interactions with surrounding environment necessary for its communication, proliferation, differentiation and migration [1][2][3][4][5][6][7][8] . A cell responds to these mechanical forces by exhibiting changes at various levels such as in morphology, protein synthesis 9,10 , gene expression 11,12 , etc. Then, cell tries to adapt to its environment through these changes. Several of its mechanical properties facilitate the adaptation of cell to the associated environment. Therefore, to understand the mechanism of working of a cell, knowledge of its mechanical properties and how do these get affected by a cell's surrounding environment is very vital. Microscopy-based imaging and its subsequent analysis is one of the traditional and widely used approaches to gain insights into the cell's mechanical and morphological properties. While microscopy imaging is a widely used approach to extract quantitative information about cell morphology, image analysis and morphology quantification are still challenging. One of the morphological features of a cell, which has been shown to exhibit several genotypic and phenotypic properties of a cell is the orientation of the cell. Similarly, orientation of fibres has been shown to exhibit various properties of associated cell population, including their density, topography, etc.The orientation of cell and the associated fibres are the indicative of many physiological and disease conditions. While microscopy-based image acquisition is the traditional method to gain insights into the orientation and morphology of cell and associated fibres, extracting thequantitative information about the c...

show abstract

Section: Global Versus Local Thresholdingmentioning

confidence: 99%

Section: Thresholdingmentioning

confidence: 99%

Section: Thresholdingmentioning

confidence: 99%

Section: Thresholdingmentioning

confidence: 99%

See 2 more Smart Citations

Extracting Cell and Fibre Orientation from Microscopic Images:Computational Methods and Tools

Sharma

2016

Current Science

View full text Add to dashboard Cite

show abstract

“… 23 – 25 It has also been integrated with the study of histological tumor sections, 26 boundary detection of epithelial cell nuclei, 27 , 28 or understanding drug influences. 29 Image-based segmentation of MSCs reported previously 30 demonstrated better results compared to conventional thresholding techniques. However, this work was more driven toward identifying all cell regions rather than individual cells, making it unsuitable for culture quality monitoring through morphological profiling of each cell.…”

Section: Introductionmentioning

confidence: 95%

Automated mesenchymal stem cell segmentation and machine learning-based phenotype classification using morphometric and textural analysis

et al. 2021

View full text Add to dashboard Cite

Purpose: Mesenchymal stem cells (MSCs) have demonstrated clinically relevant therapeutic effects for treatment of trauma and chronic diseases. The proliferative potential, immunomodulatory characteristics, and multipotentiality of MSCs in monolayer culture is reflected by their morphological phenotype. Standard techniques to evaluate culture viability are subjective, destructive, or time-consuming. We present an image analysis approach to objectively determine morphological phenotype of MSCs for prediction of culture efficacy.Approach: The algorithm was trained using phase-contrast micrographs acquired during the early and mid-logarithmic stages of MSC expansion. Cell regions are localized using edge detection, thresholding, and morphological operations, followed by cell marker identification using Hminima transform within each region to differentiate individual cells from cell clusters. Clusters are segmented using marker-controlled watershed to obtain single cells. Morphometric and textural features are extracted to classify cells based on phenotype using machine learning.Results: Algorithm performance was validated using an independent test dataset of 186 MSCs in 36 culture images. Results show 88% sensitivity and 86% precision for overall cell detection and a mean Sorensen-Dice coefficient of 0.849 AE 0.106 for segmentation per image. The algorithm exhibited an area under the curve of 0.816 (CI 95 ¼ 0.769 to 0.886) and 0.787 (CI 95 ¼ 0.716 to 0.851) for classifying MSCs according to their phenotype at early and mid-logarithmic expansion, respectively. Conclusions:The proposed method shows potential to segment and classify low and moderately dense MSCs based on phenotype with high accuracy and robustness. It enables quantifiable and consistent morphology-based quality assessment for various culture protocols to facilitate cytotherapy development.

show abstract

Automated Confluence Measurement Method for Mesenchymal Stem Cell from Brightfield Microscopic Images

Wang

Zhan

2021

Microsc Microanal

View full text Add to dashboard Cite

Cell confluence is an important metric in cell culture, as proper timing is essential to maintain cell phenotype and culture quality. To estimate cell confluence, transparent cells are observed under a phase-contrast or differential interference contrast microscope by a biologist, whose estimations are error-prone and subjective. To overcome the necessity of using the phase-contrast microscope and reducing intra- and inter-observer errors, we have proposed an algorithm that automatically measures cell confluence by using a commonly used brightfield microscope. The proposed method consists of a transport-of-intensity equation-based brightfield microscopic image processing, an image reconstruction method, and an adaptive image segmentation method based on edge detection, entropy filtering, and range filtering. Experimental results have shown that our method has outperformed several popular algorithms, with an F-score of 0.84 ± 0.07, in images with various cell confluence values. The proposed algorithm is robust and accurate enough to perform confluence measurement with various lighting conditions under a low-cost brightfield microscope, making it simple and cost-effective to use for a fully automated cell culture process.

show abstract

Image segmentation of mesenchymal stem cells in diverse culturing conditions

Cited by 3 publications

References 30 publications

Extracting Cell and Fibre Orientation from Microscopic Images:Computational Methods and Tools

Extracting Cell and Fibre Orientation from Microscopic Images:Computational Methods and Tools

Automated mesenchymal stem cell segmentation and machine learning-based phenotype classification using morphometric and textural analysis

Automated Confluence Measurement Method for Mesenchymal Stem Cell from Brightfield Microscopic Images

Contact Info

Product

Resources

About