scholarly journals Multiclass U-Net Segmentation of Brain Electron Microscopy Data

2021 ◽  
Author(s):  
Alexandra Getmanskaya ◽  
Nikolai Sokolov ◽  
Vadim Turlapov
Keyword(s):  
Electron Microscopy ◽  
State Of The Art ◽  
Training Dataset ◽  
Electron Microscopy Data ◽  
Microscopy Data ◽  
Dice Index

This work focuses on multi-class labeling and segmentation of electron microscopy data. The well-known and state-of-the-art EPFL open dataset has been labeled for 6 classes (instead of 1) and a multi-class version of the U-Net was trained. The new labeled classes are mitochondrion together with its border, mitochondrion’s border (separately), membrane, PSD, axon, vesicle. Our labeling results are available on GitHub. Our study showed that the quality of segmentation is affected by the presence of a sufficient number of specific features that distinguish the selected classes and the representation of these features in the training dataset. With 6-classes segmentation, mitochondria were segmented with the Dice index of 0.94, which is higher than with 5-classes (without mitochondrial boundaries) segmentation (Dice index of 0.892).

scholarly journals New wwPDB validation pipelines for X-ray, NMR and 3DEM structures

2014 ◽  
Vol 70 (a1) ◽  
pp. C1478-C1478
Author(s):  
Swanand Gore ◽  
Pieter Hendrickx ◽  
Eduardo Sanz-Garcia ◽  
Sameer Velankar ◽  
Gerard Kleywegt
Keyword(s):  
Electron Microscopy ◽  
Protein Data Bank ◽  
Data Bank ◽  
X Ray ◽  
Task Forces ◽  
Annotation System ◽  
Electron Microscopy Data ◽  
Microscopy Data ◽  
Machine Readable

The Protein Data Bank (PDB) is the single global archive of 3D biomacromolecular structure data. The archive is managed by the Worldwide Protein Data Bank (wwPDB; wwpdb.org) organisation through its partners, the Research Collaboratory for Structural Bioinformatics (RCSB PDB), the Protein Data Bank Japan (PDBj), the Protein Data Bank in Europe and the Biological Magnetic Resonance Bank (BMRB). Analogously, the Electron Microscopy Data Bank (EMDB) is managed by the EMDataBank (emdatabank.org) organisation. A few years ago, realising the needs and opportunities to assess the quality of biomacromolecular structures deposited in the PDB, the wwPDB and EMDataBank partners established Validation Task Forces (VTFs) to advice them on up-to-date and community-agreed methods and standards to validate X-ray, NMR and 3DEM structures and data. All three VTFs have now published their recommendations (1, 2, 3) and these are getting implemented as validation-software pipelines . The pipelines are integrated in the new joint wwPDB deposition and annotation system (http://deposit.wwpdb.org/deposition/). In addition, stand-alone servers are provided to allow practising structural biologists to validate models prior to publication and deposition (http://wwpdb.org/validation-servers.html). The validation pipelines and the output they produce (human-readable PDF reports and machine-readable XML files) will be described.

scholarly journals Label-free fluorescence predictions from large-scale correlative light and electron microscopy data

2021 ◽  
Vol 27 (S1) ◽  
pp. 94-95
Author(s):  
Ryan Lane ◽  
Luuk Balkenende ◽  
Simon van Staalduine ◽  
Anouk Wolters ◽  
Ben Giepmans ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Large Scale ◽  
Light And Electron Microscopy ◽  
Label Free ◽  
Electron Microscopy Data ◽  
Microscopy Data

scholarly journals Automatic instance segmentation of mitochondria in electron microscopy data

2021 ◽  
Author(s):  
Luke Nightingale ◽  
Joost de Folter ◽  
Helen Spiers ◽  
Amy Strange ◽  
Lucy M Collinson ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Large Scale ◽  
Machine Learning Algorithms ◽  
Public Repository ◽  
Post Processing ◽  
Human Cortex ◽  
Electron Microscopy Data ◽  
Processing Procedure ◽  
Microscopy Data ◽  
Instance Segmentation

We present a new method for rapid, automated, large-scale 3D mitochondria instance segmentation, developed in response to the ISBI 2021 MitoEM Challenge. In brief, we trained separate machine learning algorithms to predict (1) mitochondria areas and (2) mitochondria boundaries in image volumes acquired from both rat and human cortex with multi-beam scanning electron microscopy. The predictions from these algorithms were combined in a multi-step post-processing procedure, that resulted in high semantic and instance segmentation performance. All code is provided via a public repository.

scholarly journals Влияние температуры деформации в камере Бриджмена на особенности формирования дефектной структуры технически чистого титана

2018 ◽  
Vol 44 (20) ◽  
pp. 70
Author(s):  
Н.А. Шурыгина ◽  
А.М. Глезер ◽  
Д.Л. Дьяконов ◽  
А.А. Томчук ◽  
А.Г. Кадомцев ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Hydrostatic Pressure ◽  
High Hydrostatic Pressure ◽  
Defect Structures ◽  
Cryogenic Temperatures ◽  
Two Phase ◽  
Α Phase ◽  
Electron Microscopy Data ◽  
Phase Mixture ◽  
Microscopy Data

AbstractTransmission electron microscopy data showed evidence of the formation of structural regions corresponding to deformation (dislocated) fragments and dynamically recrystallized grains in α-phase titanium upon torsion at high hydrostatic pressure at room and cryogenic temperatures. It is shown that the previously proposed “two-phase mixture” model is applicable to description of these defect structures.

Sign in / Sign up

Export Citation Format

Share Document