Using simulated fluorescence cell micrographs for the evaluation of cell image segmentation algorithms

Wiesmann, Veit; Bergler, Matthias; Palmisano, Ralf; Prinzen, Martin; Franz, Daniela; Wittenberg, Thomas

doi:10.1186/s12859-017-1591-2

Cited by 20 publications

(15 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further, to localize and quantitate certain proteins within a cell, 3D evaluation of cells, including the time consuming analysis of Z-stacks, may be required. Certain circumstances allow the individual establishment of algorithms that may bypass the aforementioned difficulties ( Wiesmann et al, 2017 ). To circumvent the stated problems of image evaluation, we developed a simple and fast flow cytometry-based assay to quantitate the transport of the HTLV-1-encoded protein p8 between cells.…”

Section: Discussionmentioning

confidence: 99%

Quantitating the Transfer of the HTLV-1 p8 Protein Between T-Cells by Flow Cytometry

2018

View full text Add to dashboard Cite

The Human T-cell leukemia virus type 1 (HTLV-1)-encoded accessory protein p8 is cleaved from the precursor protein p12 encoded by the HTLV-1 open reading frame I. Both p12 and p8 are thought to contribute to efficient viral persistence. Mechanistically, p8 induces T-cell conjugates and cellular conduits. The latter are considered to facilitate transfer of p8 to target cells and virus transmission. Transfer of p8 between p8-expressing T-cells and recipient cells has been analyzed by immunofluorescence and live imaging. However, automatic quantitation of p8-transfer between cells has not been studied yet. Here we developed a novel method allowing time saving quantitation of p8 transfer between cells by flow cytometry. After establishing a protocol for the detection of intracellular p8 by flow cytometry and validation of p8 protein expression by western blot and immunofluorescence, we set up a co-culture assay between p8-expressing donor Jurkat T-cells and recipient Jurkat T-cells that had been prestained with a well-retained live cell dye. Upon quantitating the amount of p8 positive recipient cells with regard to the percentage of p8 expressing donor cells, time course experiments confirmed that p8 is rapidly transferred between Jurkat T-cells. We found that p8 enters approximately 5% of recipient T-cells immediately upon co-culture for 5 min. Prolonged co-culture for up to 24 h revealed an increase of relative p8 transfer to approximately 23% of the recipient cells. Immunofluorescence analysis of co-culture experiments and manual quantitation of p8 expression in fluorescence images confirmed the validity of the flow cytometry based assay. Application of the new assay revealed that manipulation of actin polymerization significantly decreased p8 transfer between Jurkat T-cells suggesting an important role of actin dynamics contributing to p8 transfer. Further, transfer of p8 to co-cultured T-cells varies between different donor cell types since p8 transfer could hardly been detected in co-cultures of 293T donor cells with Jurkat acceptor cells. In summary, our novel assay allows automatic and rapid quantitation of p8 transfer to target cells and might thus contribute to a better understanding of cellular processes and dynamics regulating p8 transfer and HTLV-1 transmission.

show abstract

Section: Discussionmentioning

confidence: 99%

Quantitating the Transfer of the HTLV-1 p8 Protein Between T-Cells by Flow Cytometry

2018

View full text Add to dashboard Cite

show abstract

“…Generally, there are three ways to standardize sizing of the image data. Images can be (1) distorted (shrunk or stretched), (2) masked or (3) framed 15 . We chose to standardize cropped images to a 200 by 200-pixel resolution, since most of the cell cluster segments did not exceed these dimensions.…”

Section: Resultsmentioning

confidence: 99%

Automated Classification of Bacterial Cell Sub-Populations with Convolutional Neural Networks

Tamiev

Furman

Reuel

2020

Preprint

View full text Add to dashboard Cite

Quantification of phenotypic heterogeneity present amongst bacterial cells can be a challenging task. Conventionally, classification and counting of bacteria sub-populations is achieved with manual microscopy, due to the lack of alternative, high-throughput, autonomous approaches. In this work, we apply classification-type convolutional neural networks (cCNN) to classify and enumerate bacterial cell sub-populations (B. subtilis clusters). Here, we demonstrate that the accuracy of the cCNN developed in this study can be as high as 86% when trained on a relatively small dataset (81 images). We also developed a new image preprocessing algorithm, specific to fluorescent microscope images, which increases the amount of training data available for the neural network by 72 times. By summing the classified cells together, the algorithm provides a total cell count which is on parity with manual counting, but is 10.2 times more consistent and 3.8 times faster. Finally, this work presents a complete solution framework for those wishing to learn and implement cCNN in their synthetic biology work.

show abstract

“…Determining the performance of a segmentation method requires a gold standard (or ground truth [GT]) image dataset [10,14–16]. GT datasets may be computer generated [17–19] or may consist of images that have been manually segmented by biologist specialists [15]. Recent initiatives have undertaken to make such datasets publicly available [10].…”

Section: Resultsmentioning

confidence: 99%

Nessys: A new set of tools for the automated detection of nuclei within intact tissues and dense 3D cultures

et al. 2019

View full text Add to dashboard Cite

Methods for measuring the properties of individual cells within their native 3D environment will enable a deeper understanding of embryonic development, tissue regeneration, and tumorigenesis. However, current methods for segmenting nuclei in 3D tissues are not designed for situations in which nuclei are densely packed, nonspherical, or heterogeneous in shape, size, or texture, all of which are true of many embryonic and adult tissue types as well as in many cases for cells differentiating in culture. Here, we overcome this bottleneck by devising a novel method based on labelling the nuclear envelope (NE) and automatically distinguishing individual nuclei using a tree-structured ridge-tracing method followed by shape ranking according to a trained classifier. The method is fast and makes it possible to process images that are larger than the computer’s memory. We consistently obtain accurate segmentation rates of >90%, even for challenging images such as mid-gestation embryos or 3D cultures. We provide a 3D editor and inspector for the manual curation of the segmentation results as well as a program to assess the accuracy of the segmentation. We have also generated a live reporter of the NE that can be used to track live cells in 3 dimensions over time. We use this to monitor the history of cell interactions and occurrences of neighbour exchange within cultures of pluripotent cells during differentiation. We provide these tools in an open-access user-friendly format.

show abstract

Using simulated fluorescence cell micrographs for the evaluation of cell image segmentation algorithms

Cited by 20 publications

References 24 publications

Quantitating the Transfer of the HTLV-1 p8 Protein Between T-Cells by Flow Cytometry

Quantitating the Transfer of the HTLV-1 p8 Protein Between T-Cells by Flow Cytometry

Automated Classification of Bacterial Cell Sub-Populations with Convolutional Neural Networks

Nessys: A new set of tools for the automated detection of nuclei within intact tissues and dense 3D cultures

Contact Info

Product

Resources

About