scholarly journals Robust Segmentation of Cellular Ultrastructure on Sparsely Labeled 3D Electron Microscopy Images using Deep Learning

2021 ◽  
Author(s):  
Archana Machireddy ◽  
Guillaume Thibault ◽  
Kevin G. Loftis ◽  
Kevin Stoltz ◽  
Cecilia E. Bueno ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Deep Learning ◽  
Metastatic Breast ◽  
Cellular Structures ◽  
Sem Images ◽  
Cellular Ultrastructure ◽  
Pancreatic Cancers ◽  
Robust Segmentation ◽  
3D Electron Microscopy ◽  
Tumor Biopsies

A deeper understanding of the cellular and subcellular organization of tumor cells and their interactions with the tumor microenvironment will shed light on how cancer evolves and guide effective therapy choices. Electron microscopy (EM) images can provide detailed view of the cellular ultrastructure and are being generated at an ever-increasing rate. However, the bottleneck in their analysis is the delineation of the cellular structures to enable interpretable rendering. We have mitigated this limitation by using deep learning, specifically, the ResUNet architecture, to segment cells and subcellular ultrastructure. Our initial prototype focuses on segmenting nuclei and nucleoli in 3D FIB-SEM images of tumor biopsies obtained from patients with metastatic breast and pancreatic cancers. Trained with sparse manual labels, our method results in accurate segmentation of nuclei and nucleoli with best Dice score of 0.99 and 0.98 respectively. This method can be extended to other cellular structures, enabling deeper analysis of inter- and intracellular state and interactions.

scholarly journals Robust Segmentation of Cellular Ultrastructure on Sparsely Labeled 3D Electron Microscopy Images Using Deep Learning

2021 ◽  
Author(s):  
Archana Machireddy ◽  
Guillaume Thibault ◽  
Kevin G. Loftis ◽  
Kevin Stoltz ◽  
Cecilia E. Bueno ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Deep Learning ◽  
Cellular Ultrastructure ◽  
Robust Segmentation ◽  
3D Electron Microscopy ◽  
3D Electron ◽  
Microscopy Images

scholarly journals High-quality ultrastructural preservation using cryofixation for 3D electron microscopy of genetically labeled tissues

2018 ◽  
Author(s):  
Tin Ki Tsang ◽  
Eric A. Bushong ◽  
Daniela Boassa ◽  
Junru Hu ◽  
Benedetto Romoli ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Cultured Cells ◽  
Light And Electron Microscopy ◽  
Cellular Structures ◽  
High Temporal Resolution ◽  
High Pressure Freezing ◽  
En Bloc ◽  
3D Electron Microscopy ◽  
Face Scanning ◽  
Block Face

ABSTRACTElectron microscopy (EM) offers unparalleled power to study cell substructures at the nanoscale. Cryofixation by high-pressure freezing offers optimal morphological preservation, as it captures cellular structures instantaneously in their near-native states. However, the applicability of cryofixation is limited by its incompatibilities with diaminobenzidine labeling using genetic EM tags and the high-contrast en bloc staining required for serial block-face scanning electron microscopy (SBEM). In addition, it is challenging to perform correlated light and electron microscopy (CLEM) with cryofixed samples. Consequently, these powerful methods cannot be applied to address questions requiring optimal morphological preservation and high temporal resolution. Here we developed an approach that overcomes these limitations; it enables genetically labeled, cryofixed samples to be characterized with SBEM and 3D CLEM. Our approach is broadly applicable, as demonstrated in cultured cells, Drosophila olfactory organ and mouse brain. This optimization exploits the potential of cryofixation, allowing quality ultrastructural preservation for diverse EM applications.

scholarly journals High-quality ultrastructural preservation using cryofixation for 3D electron microscopy of genetically labeled tissues

eLife ◽  
2018 ◽  
Vol 7 ◽  
Author(s):  
Tin Ki Tsang ◽  
Eric A Bushong ◽  
Daniela Boassa ◽  
Junru Hu ◽  
Benedetto Romoli ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Cultured Cells ◽  
Light And Electron Microscopy ◽  
Cellular Structures ◽  
High Pressure Freezing ◽  
Native State ◽  
En Bloc ◽  
3D Electron Microscopy ◽  
Face Scanning ◽  
Block Face

Electron microscopy (EM) offers unparalleled power to study cell substructures at the nanoscale. Cryofixation by high-pressure freezing offers optimal morphological preservation, as it captures cellular structures instantaneously in their near-native state. However, the applicability of cryofixation is limited by its incompatibility with diaminobenzidine labeling using genetic EM tags and the high-contrast en bloc staining required for serial block-face scanning electron microscopy (SBEM). In addition, it is challenging to perform correlated light and electron microscopy (CLEM) with cryofixed samples. Consequently, these powerful methods cannot be applied to address questions requiring optimal morphological preservation. Here, we developed an approach that overcomes these limitations; it enables genetically labeled, cryofixed samples to be characterized with SBEM and 3D CLEM. Our approach is broadly applicable, as demonstrated in cultured cells, Drosophila olfactory organ and mouse brain. This optimization exploits the potential of cryofixation, allowing for quality ultrastructural preservation for diverse EM applications.

scholarly journals EM-net: Deep learning for electron microscopy image segmentation

2020 ◽  
Author(s):  
Afshin Khadangi ◽  
Thomas Boudier ◽  
Vijay Rajagopal
Keyword(s):  
Electron Microscopy ◽  
Image Segmentation ◽  
Deep Learning ◽  
Focused Ion Beam ◽  
Ion Beam ◽  
Ground Truth ◽  
Cellular Structures ◽  
Low Contrast ◽  
Microscopy Image ◽  
Microscopy Techniques

AbstractRecent high-throughput electron microscopy techniques such as focused ion-beam scanning electron microscopy (FIB-SEM) provide thousands of serial sections which assist the biologists in studying sub-cellular structures at high resolution and large volume. Low contrast of such images hinder image segmentation and 3D visualisation of these datasets. With recent advances in computer vision and deep learning, such datasets can be segmented and reconstructed in 3D with greater ease and speed than with previous approaches. However, these methods still rely on thousands of ground-truth samples for training and electron microscopy datasets require significant amounts of time for carefully curated manual annotations. We address these bottlenecks with EM-net, a scalable deep convolutional neural network for EM image segmentation. We have evaluated EM-net using two datasets, one of which belongs to an ongoing competition on EM stack segmentation since 2012. We show that EM-net variants achieve better performances than current deep learning methods using small- and medium-sized ground-truth datasets. We also show that the ensemble of top EM-net base classifiers outperforms other methods across a wide variety of evaluation metrics.

scholarly journals Antitumor Potential of Green Synthesized ZnONPs Using Root Extract of Withania somnifera against Human Breast Cancer Cell Line

Separations ◽  
2021 ◽  
Vol 8 (1) ◽  
pp. 8
Author(s):  
Kollur Shiva Prasad ◽  
Shashanka K Prasad ◽  
Ravindra Veerapur ◽  
Ghada Lamraoui ◽  
Ashwini Prasad ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Electron Microscopy ◽  
Transmission Electron Microscopy ◽  
Cell Line ◽  
Withania Somnifera ◽  
Root Extract ◽  
Dependent Manner ◽  
Sem Images ◽  
Xrd Pattern ◽  
Transmission Electron

Herein we report the synthesis of zinc oxide nanoparticles (ZnONPs) using Withania somnifera root extract (WSE) as an effective chelating agent. The microscopic techniques viz., X-ray diffraction analysis (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), high-resolution transmission electron microscopy (HRTEM), and selected area electron diffraction (SAED) were employed to analyze the as-obtained ZnONPs. The crystalline planes observed from the XRD pattern agrees with the hexagonal wurtzite structure of the as-prepared ZnONPs. The aggregations and agglomerations observed in the SEM images indicated that the size of the as-prepared ZnONPs was between 30 and 43 nm. The interplanar distance between the lattice fringes observed in the HRTEM image was found to be 0.253 nm, which is in good agreement with the (100) plane obtained in the XRD pattern. Furthermore, the anti-breast cancer cytotoxic evaluation was carried out using the MCF-7 cell line, and the results showed significant cytotoxic effects in a dose-dependent manner.

scholarly journals Microscopy Methods for Biofilm Imaging: Focus on SEM and VP-SEM Pros and Cons

Biology ◽  
2021 ◽  
Vol 10 (1) ◽  
pp. 51
Author(s):  
Michela Relucenti ◽  
Giuseppe Familiari ◽  
Orlando Donfrancesco ◽  
Maurizio Taurino ◽  
Xiaobo Li ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Scanning Electron Microscopy ◽  
Ion Beam ◽  
Ruthenium Red ◽  
Variable Pressure ◽  
Sem Images ◽  
Advantages And Disadvantages ◽  
Pros And Cons ◽  
Microscopy Techniques ◽  
Scanning Electron

Several imaging methodologies have been used in biofilm studies, contributing to deepening the knowledge on their structure. This review illustrates the most widely used microscopy techniques in biofilm investigations, focusing on traditional and innovative scanning electron microscopy techniques such as scanning electron microscopy (SEM), variable pressure SEM (VP-SEM), environmental SEM (ESEM), and the more recent ambiental SEM (ASEM), ending with the cutting edge Cryo-SEM and focused ion beam SEM (FIB SEM), highlighting the pros and cons of several methods with particular emphasis on conventional SEM and VP-SEM. As each technique has its own advantages and disadvantages, the choice of the most appropriate method must be done carefully, based on the specific aim of the study. The evaluation of the drug effects on biofilm requires imaging methods that show the most detailed ultrastructural features of the biofilm. In this kind of research, the use of scanning electron microscopy with customized protocols such as osmium tetroxide (OsO4), ruthenium red (RR), tannic acid (TA) staining, and ionic liquid (IL) treatment is unrivalled for its image quality, magnification, resolution, minimal sample loss, and actual sample structure preservation. The combined use of innovative SEM protocols and 3-D image analysis software will allow for quantitative data from SEM images to be extracted; in this way, data from images of samples that have undergone different antibiofilm treatments can be compared.

scholarly journals Protocol for preparation of heterogeneous biological samples for 3D electron microscopy: a case study for insects

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Alexey A. Polilov ◽  
Anastasia A. Makarova ◽  
Song Pang ◽  
C. Shan Xu ◽  
Harald Hess
Keyword(s):  
Electron Microscopy ◽  
Biological Samples ◽  
Focused Ion Beam ◽  
Ion Beam ◽  
Three Dimensional ◽  
Entire Volume ◽  
Simple Protocol ◽  
3D Electron Microscopy ◽  
3D Em

AbstractModern morphological and structural studies are coming to a new level by incorporating the latest methods of three-dimensional electron microscopy (3D-EM). One of the key problems for the wide usage of these methods is posed by difficulties with sample preparation, since the methods work poorly with heterogeneous (consisting of tissues different in structure and in chemical composition) samples and require expensive equipment and usually much time. We have developed a simple protocol allows preparing heterogeneous biological samples suitable for 3D-EM in a laboratory that has a standard supply of equipment and reagents for electron microscopy. This protocol, combined with focused ion-beam scanning electron microscopy, makes it possible to study 3D ultrastructure of complex biological samples, e.g., whole insect heads, over their entire volume at the cellular and subcellular levels. The protocol provides new opportunities for many areas of study, including connectomics.

Sign in / Sign up

Export Citation Format

Share Document