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

Automated blood vessel segmentation is vital for biomedical imaging, asvessel changes indicate many pathologies. Still, precise segmentation isdifficult due to the complexity of vascular structures, anatomical variationsacross patients, the scarcity of annotated public datasets, and the quality ofimages. We present a thorough literature review, highlighting the state ofmachine learning techniques across diverse organs. Our goal is to provide afoundation on the topic and identify a robust baseline model for application tovascular segmentation in a new imaging modality, Hierarchical Phase ContrastTomography (HiP CT). Introduced in 2020 at the European Synchrotron RadiationFacility, HiP CT enables 3D imaging of complete organs at an unprecedentedresolution of ca. 20mm per voxel, with the capability for localized zooms inselected regions down to 1mm per voxel without sectioning. We have created atraining dataset with double annotator validated vascular data from threekidneys imaged with HiP CT in the context of the Human Organ Atlas Project.Finally, utilising the nnU Net model, we conduct experiments to assess themodels performance on both familiar and unseen samples, employing vesselspecific metrics. Our results show that while segmentations yielded reasonablyhigh scores such as clDice values ranging from 0.82 to 0.88, certain errorspersisted. Large vessels that collapsed due to the lack of hydrostatic pressure(HiP CT is an ex vivo technique) were segmented poorly. Moreover, decreasedconnectivity in finer vessels and higher segmentation errors at vesselboundaries were observed. Such errors obstruct the understanding of thestructures by interrupting vascular tree connectivity. Through our review andoutputs, we aim to set a benchmark for subsequent model evaluations usingvarious modalities, especially with the HiP CT imaging database.

Deep Learning for Vascular Segmentation and Applications in Phase Contrast Tomography Imaging