Biology subjects

Karabag, C.

Publications and source records attributed to Karabag, C..

8 recordsLinked to original sources

bioRxiv2023

Relationship Between Irregularities of the Nuclear Envelope and Mitochondria in HeLa cells Observed with Electron Microscopy

This paper describes a methodology to analyse the complexity of HeLa cells as observed with electron microscopy, in particular the relationship between mitochondria and the roughness of the nuclear envelope as reflected by the invaginations of the surface. For this purpose, several segmentation mitochondria algorithms were quantitatively compared, namely: Topology, Image Processing, Topology and Image Processing, and Deep Learning, which provided the highest accuracy. The invaginations were successfully segmented with one image processing algorithm. Metrics were extracted for both structures and correlations between the mitochondria and invaginations were explored for 25 segmented cells. It was found that there was a positive correlation between the volume of invaginations and the volume of mitochondria, and negative correlations between the number and the mean volume of mitochondria, and between the volume of the cytoplasm and the aspect ratio of mitochondria. These results suggest that there is a relationship between the shape of a cell, its nucleus and its mitochondria; as well as a relationship between the number of mitochondria and their shapes. Whilst these results were obtained from a single cell line and a relatively small number of cells, they encourage further study as the methodology proposed can be easily applied to other cells and settings. Code and data are freely available. HeLa images are available from http://dx.doi.org/10.6019/EMPIAR-10094, code from https://github.com/reyesaldasoro/MitoEM, and segmented nuclei, cells, invaginations and mitochondria from https://github.com/reyesaldasoro/HeLa_Cell_Data.

bioinformatics↗

bioRxiv2022

Evaluation of training data, ground truth and shape variability in the semantic segmentation of HeLa cells observed with electron microscopy

This paper describes the quantitative and qualitative evaluation of training data, ground truth and the variability of the shape of HeLa cells as observed with electron microscopy in the process of semantic segmentation. HeLa cells are a widely used immortal cell line, but the principles described here could apply to other cells. From a data set of 8, 000x8, 000x517 voxels, a region of interest (ROI) of 2, 000x2, 000x300 was cropped for quantitative evaluation with the qualitative evaluation performed on a larger section of interest. Patches of data and labels were used to train a U-Net architecture from scratch, which was compared also with a traditional image processing algorithm. The correctness of ground truth, that is, the inclusion of one or more nuclei within a region of interest was evaluated with a hand-drawn ground truth of the ROI. The impact of the extent of training data was evaluated by comparing results from 36,000 data and label patches extracted from the odd slices in the central region, to 135,000 patches obtained from every other slice in the set. Accuracy and Jaccard similarity index were higher for the larger training set with multiple nuclei in the ground truth. However, when the U-Nets were used to segment the section of interest beyond the ROI, results were not as good as expected, suggesting that the very complicated nature of the shapes of HeLa cells cannot be learned from a single cell. Data, code and ground truth have been made publicly released through Empair, GitHub and Zenodo.

cancer biology↗

bioRxiv2021

Volumetric Semantic Instance Segmentation of the Plasma Membrane of HeLa Cells

In this work, the unsupervised volumetric semantic segmentation of the plasma membrane of HeLa cells as observed with Serial Block Face Scanning Electron Microscopy is described. The resin background of the images was segmented at different slices of a 3D stack of 518 slices with 8, 192 x 8, 192 pixels each. The background was used to create a distance map which helped identify and rank the cells by their size at each slice. The centroids of the cells detected at different slices were linked to identify them as a single cell that spanned a number of slices. A subset of these cells, i.e., largest ones and those not close to the edges were selected for further processing. The selected cells were then automatically cropped to smaller regions of interest of 2, 000 x 2, 000 x 300 voxels that were treated as cell instances. Then, for each of these volumes the nucleus was segmented and the cell was separated from any neighbouring cells through a series of traditional image processing steps that followed the plasma membrane. The segmentation process was repeated for all the regions selected. For one cell for which the ground truth was available, the algorithm provided excellent results in Accuracy (AC) and Jaccard Index (JI): Nucleus: JI = 0.9665, AC= 0.9975, Cell and Nucleus JI = 0.8711, AC = 0.9655, Cell only JI = 0.8094, AC = 0.9629. A limitation of the algorithm for the plasma membrane segmentation was the presence of background, as in cases of tightly packed cells. When tested for these conditions, the segmentation of the nuclear envelope was still possible. All the code and data are released openly through GitHub, Zenodo and EMPIAR.

cell biology↗

bioRxiv2020

Estimation of cellularity in tumours treated with Neoadjuvant therapy: A comparison of Machine Learning algorithms

This paper describes a method for residual tumour cellularity (TC) estimation in Neoadjuvant treatment (NAT) of advanced breast cancer. This is determined manually by visual inspection by a radiologist, then an automated computation will contribute to reduce time workload and increase precision and accuracy. TC is estimated as the ratio of tumour area by total image area estimated after the NAT. The method proposed computes TC by using machine learning techniques trained with information on morphological parameters of segmented nuclei in order to classify regions of the image as tumour or normal. The data is provided by the 2019 SPIE Breast challenge, which was proposed to develop automated TC computation algorithms. Three algorithms were implemented: Support Vector Machines, Nearest K-means and Adaptive Boosting (AdaBoost) decision trees. Performance based on accuracy is compared and evaluated and the best result was obtained with Support Vector Machines. Results obtained by the methods implemented were submitted during ongoing challenge with a maximum of 0.76 of prediction probability of success.

pathology↗

bioRxiv2020

Morphological estimation of Cellularity on Neo-adjuvant treated breast cancer histological images

This paper describes a methodology that extracts morphological features from histological breast cancer images stained for Hematoxilyn and Eosin (H&E). Cellularity was estimated and the correlation between features and the residual tumour size cellularity after a Neo-Adjuvant treatment (NAT) was examined. Images from whole slide imaging (WSI) were processed automatically with traditional computer vision methods to extract twenty two morphological parameters from the nuclei, epithelial region and the global image. The methodology was applied to a set of images from breast cancer under NAT. The data came from the BreastPathQ Cancer Cellularity Challenge 2019, and consisted of 2579 patches of 255x255 pixels of H&E histopatological samples from NAT treatment patients. The methodology automatically implements colour separation, segmentation and morphological analysis using traditional algorithms (K-means grouping, watershed segmentation, Otsus binarisation). Linear regression methods were applied to determine strongest correlation between the parameters and the cancer cellularity. The morphological parameters showed correlation with the residual tumour cancer cellularity. The strongest correlations corresponded to the stroma concentration value (r = -0.9786) and value from HSV image colour space (r = -0.9728), both from a global image parameters.

pathology↗

bioRxiv2020

Geometric differences between nuclear envelopes of Wild-type and Chlamydia trachomatis-infected HeLa cells

In this work, the geometrical characteristics of two different types of cells observed with Electron Microscopy were analysed. The nuclear envelope of Wild-type HeLa cells and Chlamydia trachomatis-infected HeLa cells were automatically segmented and then modelled against a spheroid and converted to a two-dimensional surface. Geometric measurements from this surface and the volumetric nuclear envelope were extracted to compare the two types of cells. The measurements included the nuclear volume, the sphericity of the nucleus, its flatness or spikiness. In total 13 different cells were segmented: 7 Wild-type and 6 Chlamydia trachomatis-infected. The cells were statistically different in the following measurements. Wild-type HeLa cells have greater volumes than that of Chlamydia trachomatis-infected HeLa cells and they are more spherical as Jaccard index suggests. Standard deviation ({sigma}), and range of values for the nuclear envelope, which shows the distance of the highest peaks and deepest valleys from the spheroid, were also extracted from the modelling against a spheroid and these metrics were used to compare two different data sets in order to draw conclusions.

cell biology↗

bioRxiv2020

Semantic Segmentation of HeLa Cells: An Objective Comparison between one Traditional Algorithm and Three Deep-Learning Architectures

In this work, images of a HeLa cancer cell were semantically segmented with one traditional image-processing algorithm and three deep learning architectures: VGG16, ResNet18 and Inception-ResNet-v2. Three hundred slices, each 2000 x 2000 pixels, of a HeLa Cell were acquired with Serial Block Face Scanning Electron Microscopy. The deep learning architectures were pre-trained with ImageNet and then fine-tuned with transfer learning. The image-processing algorithm followed a pipeline of several traditional steps like edge detection, dilation and morphological operators. The algorithms were compared by measuring pixel-based segmentation accuracy and Jaccard index against a labelled ground truth. The results indicated a superior performance of the traditional algorithm (Accuracy = 99%, Jaccard = 93%) over the deep learning architectures: VGG16 (93%, 90%), ResNet18 (94%, 88%), Inception-ResNet-v2 (94%, 89%).

cancer biology↗

bioRxiv2018

Modelling the nuclear envelope of HeLa cells

This paper describes a framework for the automatic segmentation of the nuclear envelope of cancerous HeLa cells and the modelling of the volumetric shape against an ellipsoid. The framework is automatic and unsupervised and reported a Jaccard Similarity Index of 0.968 against a manual segmentation. The modelling of the surface provides a visual display of the variations, both smooth and rugged over the surface, and measurements can be extracted with the expectation that they can correlate with the biological characteristics of the cells.

cancer biology↗