Multiclass semantic segmentation and quantification of traumatic brain injury lesions on head CT using deep learning: an algorithm development and multicentre validation study.

Journal: The Lancet. Digital health
Published Date: May 14, 2020

Abstract

BACKGROUND: CT is the most common imaging modality in traumatic brain injury (TBI). However, its conventional use requires expert clinical interpretation and does not provide detailed quantitative outputs, which may have prognostic importance. We aimed to use deep learning to reliably and efficiently quantify and detect different lesion types.

Authors

  • Miguel Monteiro
  • Virginia F J Newcombe
    University Division of Anaesthesia, Department of Medicine, Cambridge University, UK; Wolfson Brain Imaging Centre, Cambridge University, UK.
  • Francois Mathieu
    Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, UK.
  • Krishma Adatia
    Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, UK.
  • Konstantinos Kamnitsas
    Biomedical Image Analysis Group, Imperial College London, UK. Electronic address: konstantinos.kamnitsas12@imperial.ac.uk.
  • Enzo Ferrante
  • Tilak Das
    Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, UK.
  • Daniel Whitehouse
    Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, UK.
  • Daniel Rueckert
    Biomedical Image Analysis (BioMedIA) Group, Department of Computing, Imperial College London, UK. Electronic address: d.rueckert@imperial.ac.uk.
  • David K Menon
    University Division of Anaesthesia, Department of Medicine, Cambridge University, UK; Wolfson Brain Imaging Centre, Cambridge University, UK.
  • Ben Glocker
    Kheiron Medical Technologies, London, UK.

Keywords

External Resources

Popular Topics
Recent Journals