scholarly journals Self-Supervised Pretraining for Transferable Quantitative Phase Image Cell Segmentation

2021 ◽  
Author(s):  
Tomas Vicar ◽  
Jiri Chmelik ◽  
Roman Jakubicek ◽  
Larisa Chmelikova ◽  
Jaromir Gumulec ◽  
...  
Keyword(s):  
Cell Segmentation ◽  
Phase Image ◽  
Quantitative Phase

scholarly journals Self-Supervised Pretraining for Transferable Quantitative Phase Image Cell Segmentation

2021 ◽  
Author(s):  
Tomas Vicar ◽  
Jiri Chmelik ◽  
Roman Jakubicek ◽  
Larisa Chmelikova ◽  
Jaromir Gumulec ◽  
...  
Keyword(s):  
Cell Types ◽  
Cell Segmentation ◽  
Phase Image ◽  
Adherent Cell ◽  
Post Processing ◽  
Quantitative Phase ◽  
Microscopy Image ◽  
Processing Step ◽  
Unlabelled Data ◽  
Different Cell Types

In this paper, U-Net-based method for robust adherent cell segmentation for quantitative phase microscopy image is designed and optimised. We designed and evaluated four specific post-processing pipelines. To increase the transferability to different cell types, non-deep learning transfer with adjustable parameters is used in the post-processing step. Additionally, we proposed a self-supervised pretraining technique using nonlabelled data, which is trained to reconstruct multiple image distortions and improved the segmentation performance by from 0.67 to 0.70 of Object-wise Intersection over Union. Moreover, we publish a new dataset of manually labelled images suitable for this task together with the unlabelled data for self-supervised pretraining.

scholarly journals Tissue spatial correlation as cancer marker

2018 ◽  
Author(s):  
Masanori Takabayashi ◽  
Hassaan Majeed ◽  
Andre Kajdacsy-Balla ◽  
Gabriel Popescu
Keyword(s):  
Spatial Autocorrelation ◽  
Fourier Transforms ◽  
Interference Microscopy ◽  
Phase Image ◽  
Phase Imaging ◽  
Imaging Data ◽  
Cancer Marker ◽  
Fixed Tissue ◽  
Quantitative Phase ◽  
Autocorrelation Length

AbstractWe propose a new intrinsic cancer marker in fixed tissue biopsy slides, which is based on the local spatial autocorrelation length obtained from quantitative phase images. The spatial autocorrelation length in a small region of the tissue phase image is sensitive to the nanoscale cellular morphological alterations and can hence inform on carcinogenesis. Therefore, this metric can potentially be used as an intrinsic cancer marker in histopathology. Typically, these correlation length maps are calculated by computing 2D Fourier transforms over image sub-regions – requiring long computational times. In this paper, we propose a more time efficient method of computing the correlation map and demonstrate its value for diagnosis of benign and malignant breast tissues. Our methodology is based on highly sensitive quantitative phase imaging data obtained by spatial light interference microscopy (SLIM).

Modeling quantitative phase image formation under tilted illuminations

2012 ◽  
Vol 37 (10) ◽  
pp. 1718 ◽  
Author(s):  
Pierre Bon ◽  
Benoit Wattellier ◽  
Serge Monneret
Keyword(s):  
Image Formation ◽  
Phase Image ◽  
Quantitative Phase

scholarly journals Automated Cell Segmentation for Quantitative Phase Microscopy

2018 ◽  
Vol 37 (4) ◽  
pp. 929-940 ◽  
Author(s):  
Nathan O. Loewke ◽  
Sunil Pai ◽  
Christine Cordeiro ◽  
Dylan Black ◽  
Bonnie L. King ◽  
...  
Keyword(s):  
Cell Segmentation ◽  
Quantitative Phase ◽  
Phase Microscopy ◽  
Quantitative Phase Microscopy

scholarly journals Morphological Measurement of Living Cells in Methanol with Digital Holographic Microscopy

2013 ◽  
Vol 2013 ◽  
pp. 1-7 ◽  
Author(s):  
Yunxin Wang ◽  
Yishu Yang ◽  
Dayong Wang ◽  
Liting Ouyang ◽  
Yizhuo Zhang ◽  
...  
Keyword(s):  
Morphological Change ◽  
Optical Thickness ◽  
Hela Cells ◽  
Living Cells ◽  
Digital Holographic Microscopy ◽  
Phase Image ◽  
Toxicity Screening ◽  
Label Free ◽  
Quantitative Phase ◽  
Holographic Microscopy

Cell morphology is the research foundation in many applications related to the estimation of cell status, drug response, and toxicity screening. In biomedical field, the quantitative phase detection is an inevitable trend for living cells. In this paper, the morphological change of HeLa cells treated with methanol of different concentrations is detected using digital holographic microscopy. The compact image-plane digital holographic system is designed based on fiber elements. The quantitative phase image of living cells is obtained in combination with numerical analysis. The statistical analysis shows that the area and average optical thickness of HeLa cells treated with 12.5% or 25% methanol reduce significantly, which indicates that the methanol with lower concentration could cause cellular shrinkage. The area of HeLa cells treated with 50% methanol is similar to that of normal cells(P>0.05), which reveals the fixative effect of methanol with higher concentration. The maximum optical thickness of the cells treated with 12.5%, 25%, and 50% methanol is greater than that of untreated cells, which implies the pyknosis of HeLa cells under the effect of methanol. All of the results demonstrate that digital holographic microscopy has supplied a noninvasive imaging alternative to measure the morphological change of label-free living cells.

Sign in / Sign up

Export Citation Format

Share Document