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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Three-dimensional (3D) freehand ultrasound (US) reconstruction without usingany additional external tracking device has seen recent advances with deepneural networks (DNNs). In this paper, we first investigated two identifiedcontributing factors of the learned inter-frame correlation that enable theDNN-based reconstruction: anatomy and protocol. We propose to incorporate theability to represent these two factors - readily available during training - asthe privileged information to improve existing DNN-based methods. This isimplemented in a new multi-task method, where the anatomical and protocoldiscrimination are used as auxiliary tasks. We further develop a differentiablenetwork architecture to optimise the branching location of these auxiliarytasks, which controls the ratio between shared and task-specific networkparameters, for maximising the benefits from the two auxiliary tasks.Experimental results, on a dataset with 38 forearms of 19 volunteers acquiredwith 6 different scanning protocols, show that 1) both anatomical and protocolvariances are enabling factors for DNN-based US reconstruction; 2) learning howto discriminate different subjects (anatomical variance) and predefined typesof scanning paths (protocol variance) both significantly improve frameprediction accuracy, volume reconstruction overlap, accumulated tracking errorand final drift, using the proposed algorithm.

Privileged Anatomical and Protocol Discrimination in Trackerless 3D Ultrasound Reconstruction