Cell Segmentation for Image Cytometry: Advances, Insufficiencies, and Challenges

Wang, Zhenzhou

doi:10.1002/cyto.a.23686

Cited by 27 publications

(22 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In these cases, despite the real-time constraints, this is only possible with a manual removal of the artifacts or with non-real-time traditional algorithms. As it is shown in [44], [45], [46], the non-trainable segmentation algorithms need a parameter optimization, which depends on image quality and cell clusters. Furthermore, they need a cell identification to distinguish artifacts.…”

Section: Experimental Results and Robustness Analysismentioning

confidence: 99%

Neuroblastoma Cells Classification Through Learning Approaches by Direct Analysis of Digital Holograms

Priscoli

Memmolo

Ciaparrone

et al. 2021

IEEE J. Select. Topics Quantum Electron.

View full text Add to dashboard Cite

The label-free single cell analysis by machine and Deep Learning, in combination with digital holography in transmission microscope configuration, is becoming a powerful framework exploited for phenotyping biological samples. Usually, quantitative phase images of cells are retrieved from the reconstructed complex diffraction patterns and used as inputs of a deep neural network. However, the phase retrieval process can be very time consuming and prone to errors. Here we address the classification of cells by using learning strategies with images coming directly from the raw recorded digital holograms, i.e. without any data processing or refocusing involved. Indeed, in the raw digital hologram the entire complex amplitude information of the sample is intrinsically embedded in the form of modulated fringes. We develop a training strategy, based on deep and feature based machine learning models, in order extract such information by skipping the classical reconstruction process for classifying different neuroblastoma cells. We provided an experimental validation by using the proposed strategy to classify two neuroblastoma cell lines.

show abstract

Section: Experimental Results and Robustness Analysismentioning

confidence: 99%

Neuroblastoma Cells Classification Through Learning Approaches by Direct Analysis of Digital Holograms

Priscoli

Memmolo

Ciaparrone

et al. 2021

IEEE J. Select. Topics Quantum Electron.

View full text Add to dashboard Cite

show abstract

“…Thresholding and clustering methods suffer from several limitations when pixel intensity histograms are not well equalized or in a situation of elevated noise. Slope difference distribution such as in Wang’s method 9 overcome some of these issues although still lacking a full unsupervised implementation 10 . A considerable improvement over cell segmentation based on direct thresholding of the acquired image, has been the introduction of machine-learning approaches, such as that implemented in ilastik 11 , that permits to feed to the segmentation algorithm one metric that is an estimation, based on the information conveyed by multiple images or layers, of the probability of a pixel to belong to a certain category or cluster, or the use of deep-learning algoritms 12 , topic that will be not expanded in this introduction because it is beyond the purpose of the present article.…”

Section: Introductionmentioning

confidence: 99%

RUNIMC: An R-based package for imaging mass cytometry data analysis and pipeline validation

Dolcetti

Barber

Weitsman

et al. 2021

Preprint

View full text Add to dashboard Cite

We propose a novel pipeline for the analysis of imaging mass cytometry data, comparing an unbiased approach, representing the actual gold standard, with a novel biased method. We made use of both synthetic/ controlled datasets as well as two datasets obtained from FFPE sections of follicular lymphoma, and head and neck patients, stained with a 14 and 29-markers panels respectively. The novel pipeline, denominated RUNIMC, has been completely developed in R and contained in a single package. The novelty resides in the ease with which multi-class random forest classifier can be used to classify image features, making the pathologist and expert classification pivotal, and the use of a random forest regression approach that permits a better detection of cell boundaries, and alleviates the necessity of relying on a perfect nuclear staining.

show abstract

“…DAPI) to identify putative cell centers 7,9,16 . Unfortunately, even such one-channel segmentation is challenging, commonly requiring manual tuning and corrections 18 , including compensation for physical misalignment of molecular and auxiliary stains. The nuclei positions also do not inform on the extent of the cell body.…”

mentioning

confidence: 99%

“…Existing cell segmentation methods rely on nuclear (DAPI) or cytoplasmic (poly-A) staining 8,9,15 , segmenting the images with watershed or other algorithms to obtain cell labels 18,23 . While Baysor can perform segmentation using only the information on the measured molecules ( Fig.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Bayesian segmentation of spatially resolved transcriptomics data

Petukhov

Khodosevich

Soldatov

et al. 2020

Preprint

View full text Add to dashboard Cite

Spatial transcriptomics is an emerging stack of technologies, which adds spatial dimension to conventional single-cell RNA-sequencing. New protocols, based on in situ sequencing or multiplexed RNA fluorescent in situ hybridization register positions of single molecules in fixed tissue slices. Analysis of such data at the level of individual cells, however, requires accurate identification of cell boundaries. While many existing methods are able to approximate cell center positions using nuclei stains, current protocols do not report robust signal on the cell membranes, making accurate cell segmentation a key barrier for downstream analysis and interpretation of the data. To address this challenge, we developed a tool for Bayesian Segmentation of Spatial Transcriptomics Data (Baysor), which optimizes segmentation considering the likelihood of transcriptional composition, size and shape of the cell. The Bayesian approach can take into account nuclear or cytoplasm staining, however can also perform segmentation based on the detected transcripts alone. We show that Baysor segmentation can in some cases nearly double the number of the identified cells, while reducing contamination. Importantly, we demonstrate that Baysor performs well on data acquired using five different spatially-resolved protocols, making it a useful general tool for analysis of high-resolution spatial data.

show abstract

Cell Segmentation for Image Cytometry: Advances, Insufficiencies, and Challenges

Cited by 27 publications

References 14 publications

Neuroblastoma Cells Classification Through Learning Approaches by Direct Analysis of Digital Holograms

Neuroblastoma Cells Classification Through Learning Approaches by Direct Analysis of Digital Holograms

RUNIMC: An R-based package for imaging mass cytometry data analysis and pipeline validation

Bayesian segmentation of spatially resolved transcriptomics data

Contact Info

Product

Resources

About