English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Electron microscopy (EM) allows the identification of intracellularorganelles such as mitochondria, providing insights for clinical and scientificstudies. In recent years, a number of novel deep learning architectures havebeen published reporting superior performance, or even human-level accuracy,compared to previous approaches on public mitochondria segmentation datasets.Unfortunately, many of these publications do not make neither the code nor thefull training details public to support the results obtained, leading toreproducibility issues and dubious model comparisons. For that reason, andfollowing a recent code of best practices for reporting experimental results,we present an extensive study of the state-of-the-art deep learningarchitectures for the segmentation of mitochondria on EM volumes, and evaluatethe impact in performance of different variations of 2D and 3D U-Net-likemodels for this task. To better understand the contribution of each component,a common set of pre- and post-processing operations has been implemented andtested with each approach. Moreover, an exhaustive sweep of hyperparametersvalues for all architectures have been performed and each configuration hasbeen run multiple times to report the mean and standard deviation values of theevaluation metrics. Using this methodology, we found very stable architecturesand hyperparameter configurations that consistently obtain state-of-the-artresults in the well-known EPFL Hippocampus mitochondria segmentation dataset.Furthermore, we have benchmarked our proposed models on two other availabledatasets, Lucchi++ and Kasthuri++, where they outperform all previous works.The code derived from this research and its documentation are publiclyavailable.

Stable deep neural network architectures for mitochondria segmentation on electron microscopy volumes