B
Ben Glocker
Researcher at Imperial College London
Publications - 363
Citations - 30047
Ben Glocker is an academic researcher from Imperial College London. The author has contributed to research in topics: Computer science & Segmentation. The author has an hindex of 60, co-authored 300 publications receiving 20402 citations. Previous affiliations of Ben Glocker include Analysis Group & Microsoft.
Papers
More filters
Book ChapterDOI
Computing CNN Loss and Gradients for Pose Estimation with Riemannian Geometry
Benjamin Hou,Nina Miolane,Bishesh Khanal,Bishesh Khanal,Matthew C. H. Lee,Amir Alansary,Steven McDonagh,Joseph V. Hajnal,Daniel Rueckert,Ben Glocker,Bernhard Kainz +10 more
TL;DR: This work proposes a new metric for linear spaces that does not take into account the Lie group structure of SE(3) that is applicable to pose estimation and regression.
Proceedings ArticleDOI
Real-time respiratory motion tracking: roadmap correction for hepatic artery catheterizations
Selen Atasoy,Martin Groher,Darko Zikic,Ben Glocker,Tobias Waggershauser,Marcus Dr. Pfister,Nassir Navab +6 more
TL;DR: The objective of this work is to introduce dynamic roadmaps into clinical workflow for hepatic artery catheterizations and allow for continuous visualization of the vessels in 2D fluoroscopy images without additional contrast injection.
Machine Learning with Multi-Site Imaging Data: An Empirical Study on the Impact of Scanner Effects.
TL;DR: In this paper, structural T1-weighted brain MRI obtained from two different studies, Cam-CAN and UK Biobank, was used to investigate the impact of scanner effects when using machine learning on multi-site neuroimaging data.
Book ChapterDOI
Neighbourhood approximation forests.
TL;DR: "neighbourhood approximation forests" (NAF) is a supervised learning algorithm that approximates the neighbourhood structure resulting from an arbitrary distance that can be applied to distances based on ground truth annotations.
Book ChapterDOI
Robust registration of longitudinal spine CT
TL;DR: Quantitative evaluation on a database of 93 patients with a total of 276 registrations on longitudinal spine CT demonstrate that the registration approach which incorporates estimates of vertebrae locations obtained from a learning-based classification method significantly reduces the number of failure cases.