Mark A Pinnock1,2, Yipeng Hu1,2, Alan Bainbridge3, David Atkinson4, Rajeshwar P Mookerjee5, Stuart A Taylor4, Dean C Barratt1,2, and Manil D Chouhan4
1Centre for Medical Image Computing, University College London, London, United Kingdom, 2Wellcome/EPSRC Centre for Interventional and Surgical Sciences, University College London, London, United Kingdom, 3Department of Medical Physics and Biomedical Engineering, University College London Hospitals NHS Foundation Trust, London, United Kingdom, 4Centre for Medical Imaging, Division of Medicine, University College London, London, United Kingdom, 5Institute for Liver and Digestive Health, Division of Medicine, University College London, London, United Kingdom
1Centre for Medical Image Computing, University College London, London, United Kingdom, 2Wellcome/EPSRC Centre for Interventional and Surgical Sciences, University College London, London, United Kingdom, 3Department of Medical Physics and Biomedical Engineering, University College London Hospitals NHS Foundation Trust, London, United Kingdom, 4Centre for Medical Imaging, Division of Medicine, University College London, London, United Kingdom, 5Institute for Liver and Digestive Health, Division of Medicine, University College London, London, United Kingdom
Deep learning-based detection of liver disease from
T2-weighted MRI sequences/relevant organ segmentation mask only is feasible.
Figure 1: GradCAM visualisations for two false positive classifications
Figure 2: GradCAM visualisations for two true positive classifications