Localisation of Vertebrae on DXA Images using Constrained Local Models with Random Forest Regression Voting

Paul Bromiley, Timothy Cootes, Jianhua Yao (Editor), Ben Glocker (Editor), Tobias Klinder (Editor), Shuo Li (Editor)

    Research output: Chapter in Book/Conference proceedingConference contributionpeer-review

    195 Downloads (Pure)

    Abstract

    Fractures associated with osteoporosis are a significant public health risk, and one that is likely to increase with an ageing population. However, many osteoporotic vertebral fractures present on images do not come to clinical attention or lead to preventative treatment. Furthermore, vertebral fracture assessment (VFA) typically depends on subjective judgement by a radiologist. The potential utility of computer-aided VFA systems is therefore considerable. Previous work has shown that Active Appearance Models (AAMs) give accurate results when locating landmarks on vertebra in DXA images, but can give poor fits in a substantial subset of examples, particularly the more severe fractures. Here we evaluate Random Forest Regression Voting Constrained Local Models (RFRV-CLMs) for this task and show that, while they lead to slightly poorer median errors than AAMs, they are much more robust, reducing the proportion of fit failures by 68\%. They are thus more suitable for use in computer-aided VFA systems.
    Original languageEnglish
    Title of host publicationRecent Advances in Computational Methods and Clinical Applications for Spine Imaging
    EditorsJianhua Yao, Ben Glocker, Tobias Klinder, Shuo Li
    Place of PublicationSwitzerland
    PublisherSpringer Nature
    Pages159-171
    Number of pages13
    Volume20
    ISBN (Print)978-3-319-14147-3
    DOIs
    Publication statusPublished - 14 Sept 2014
    EventMICCAI Workshop on Computational Methods and Clinical Applications for Spine Imaging (CSI 2014) - Boston, USA
    Duration: 14 Sept 201414 Sept 2014

    Publication series

    NameLecture Notes in Computational Vision and Biomechanics

    Conference

    ConferenceMICCAI Workshop on Computational Methods and Clinical Applications for Spine Imaging (CSI 2014)
    CityBoston, USA
    Period14/09/1414/09/14

    Fingerprint

    Dive into the research topics of 'Localisation of Vertebrae on DXA Images using Constrained Local Models with Random Forest Regression Voting'. Together they form a unique fingerprint.

    Cite this