Detection and segmentation of morphologically complex eukaryotic cells in fluorescence microscopy images via feature pyramid fusion

Korfhage, Nikolaus; Mühling, Markus; Ringshandl, Stephan; Becker, Anke; Schmeck, Bernd; Freisleben, Bernd

doi:10.1371/journal.pcbi.1008179

Cited by 28 publications

(22 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A variety of neural net and other architectures have been explored to improve detection and classification accuracy and expand generalizability to new types of images. Several deep learning architectures developed for natural images have been adapted for marker detection in images of cells including Fully Convolutional Networks (FCNs) (Lux and Matula, 2020), Visual Geometry Group (VGG16) (Wang et al, 2019;Shahzad M et al, 2020), Residual Networks (ResNets) (Lee and Jeong, 2020), UNet (Al-Kofahi et al, 2018;McQuin et al, 2018;Schmidt et al, 2018;Wen et al, 2018;Vu et al, 2019;Horwath et al, 2020;Lugagne, Lin and Dunlop, 2020), and Mask R-CNN (Kromp et al, 2019;Vuola, Akram andKannala, 2019, 2019;Korfhage et al, 2020;Liu et al, 2020;Masubuchi et al, 2020). In classical image analysis, advances in methodology commonly involve the development of new algorithms; any changes in parameter settings needed to accommodate new data are regarded as project-specific details.…”

Section: Related Workmentioning

confidence: 99%

UnMICST: Deep learning with real augmentation for robust segmentation of highly multiplexed images of human tissues

Yapp

Novikov

Jang

et al. 2021

Preprint

View full text Add to dashboard Cite

Newly developed technologies have made it feasible to routinely collect highly multiplexed (20-60 channel) images at subcellular resolution from human tissues for research and diagnostic purposes. Extracting single cell data from such images requires efficient and accurate image segmentation. This starts with identification of nuclei, a challenging problem in tissue imaging that has recently benefited from deep learning. In this paper we demonstrate two generally applicable approaches to improving segmentation accuracy as scored using new human-labelled segmentation masks spanning multiple human tissues. The first approach involves the use of real augmentations during training. These comprise defocused and saturated image data and improve model accuracy when computational augmentation (Gaussian blurring) does not. The second involves collection of nuclear envelope data. The two approaches cumulatively and substantially improve segmentation with three different deep learning frameworks, yielding a set of high accuracy segmentation models. Moreover, the use of real augmentation may have applications outside of microscopy.

show abstract

Section: Related Workmentioning

confidence: 99%

UnMICST: Deep learning with real augmentation for robust segmentation of highly multiplexed images of human tissues

Yapp

Novikov

Jang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…It is quite challenging as there are high chances of errors in separating each cell. Specially designed deep learning models for segmenting densely packed cell regions are reported in works like (Korfhage et al, 2020) . It uses an object detection module called a feature pyramid network, which apparently outperforms MRCNN in this task.…”

Section: Acknowledgementmentioning

confidence: 99%

“…(Payer, Štern, Feiner, Bischof, & Urschler, 2019) where the pipeline makes predictions for every cell instance in videos, as well produces temporally connected instance segmentations. Cell instance segmentation in calcium imaging videos is described in (Kirschbaum, Bailoni, & Hamprecht, 2020) (Korfhage et al, 2020) . It uses an object detection module called a feature pyramid network, which apparently outperforms MRCNN in this task.…”

Section: Simulation Of Image Artefactsmentioning

confidence: 99%

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

Kar

Petit

Refahi

et al. 2021

Preprint

View full text Add to dashboard Cite

Segmenting three dimensional microscopy images is essential for understanding phenomena like morphogenesis, cell division, cellular growth and genetic expression patterns. Recently, deep learning (DL) pipelines have been developed which claim to provide high accuracy segmentation of cellular images and are increasingly considered as the state-of-the-art for image segmentation problems. However, it remains difficult to define their relative performance as the concurrent diversity and lack of uniform evaluation strategies makes it difficult to know how their results compare. In this paper, we first made an inventory of the available DL methods for 3D segmentation. We next implemented and quantitatively compared a number of representative DL pipelines, alongside a highly efficient non-DL method named MARS. The DL methods were trained on a common dataset of 3D cellular confocal microscopy images. Their segmentation accuracies were also tested in the presence of different image artefacts. A new method for segmentation quality evaluation was adopted which isolates segmentation errors due to under/over segmentation. This is complemented with new visualisation strategies that make interactive exploration of segmentation quality possible. Our analysis shows that the DL pipelines have very different levels of accuracy. Two of them show high performance, and offer clear advantages in terms of adaptability to new data.

show abstract

“…Most cell counting tools developed through the usage of DL are based on recurring object segmentation in frameworks such as Mask R-CNN [ 12 , 13 ], U-NET [ 14 – 16 ], or feature pyramid networks [ 17 ]. However, this type of approach demands pixel-wise labeling of the ground truth, which usually is difficult to obtain.…”

Section: Introductionmentioning

confidence: 99%

Object detection for automatic cancer cell counting in zebrafish xenografts

et al. 2021

View full text Add to dashboard Cite

Cell counting is a frequent task in medical research studies. However, it is often performed manually; thus, it is time-consuming and prone to human error. Even so, cell counting automation can be challenging to achieve, especially when dealing with crowded scenes and overlapping cells, assuming different shapes and sizes. In this paper, we introduce a deep learning-based cell detection and quantification methodology to automate the cell counting process in the zebrafish xenograft cancer model, an innovative technique for studying tumor biology and for personalizing medicine. First, we implemented a fine-tuned architecture based on the Faster R-CNN using the Inception ResNet V2 feature extractor. Second, we performed several adjustments to optimize the process, paying attention to constraints such as the presence of overlapped cells, the high number of objects to detect, the heterogeneity of the cells’ size and shape, and the small size of the data set. This method resulted in a median error of approximately 1% of the total number of cell units. These results demonstrate the potential of our novel approach for quantifying cells in poorly labeled images. Compared to traditional Faster R-CNN, our method improved the average precision from 71% to 85% on the studied data set.

show abstract

Detection and segmentation of morphologically complex eukaryotic cells in fluorescence microscopy images via feature pyramid fusion

Cited by 28 publications

References 37 publications

UnMICST: Deep learning with real augmentation for robust segmentation of highly multiplexed images of human tissues

UnMICST: Deep learning with real augmentation for robust segmentation of highly multiplexed images of human tissues

Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

Object detection for automatic cancer cell counting in zebrafish xenografts

Contact Info

Product

Resources

About