Towards single-stage prediction of vertebrae segmentation for opportunistic screening of 3D CT scans

Hellena Hempe, Mattias P Heinrich

Keywords: Spine CT, Semantic segmentation, Localisation, Light-weight, Efficiency

Abstract:

Opportunistic screening of the spine using routine CT calls for light-weight localisation and segmentation networks.
We explore the performance of a single-stage network based on the MobileNetV3 and the DeepLab/ASPP on the VerSe20 Challenge dataset.
Our proposed method improves the robustness of localisation and segmentation in comparison to other design choices of network architectures.

Paper Type: both

Primary Subject Area: Segmentation

Secondary Subject Area: Application: Radiology

Paper Status: original work, not submitted yet

Source Code Url: https://github.com/hehempe/mobile-aspp-verse

Data Set Url: https://osf.io/4skx2/

Registration: I acknowledge that publication of this at MIDL and in the proceedings requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.