Abstract #1696

Assessment of Automated Brain Tumour Segmentation Tools for Clinical Data

Reneira Seeamber¹, Katherine L Ordidge^2,3, Felice d’Arco⁴, Kshitij Mankad⁴, Tara D Barwick^2,3, Adam D Waldman^5,6, Patrick W Hales⁷, and Matthew Grech-Sollars^2,3

¹Department of Computing, Imperial College London, London, United Kingdom, ²Department of Surgery and Cancer, Imperial College London, London, United Kingdom, ³Department of Imaging, Imperial College Healthcare NHS Trust UK, London, United Kingdom, ⁴Great Ormond Street Children’s Hospital, London, United Kingdom, ⁵Department of Medicine, Imperial College London, London, United Kingdom, ⁶Centre for Clinical Brain Sciences, The University of Edinburgh, Edinburgh, United Kingdom, ⁷Department of Developmental Neurosciences, UCL Great Ormond Street Institute of Child Health, London, United Kingdom

Radiological diagnosis of certain brain tumours remains poor. The development of automated tools that can extract valuable, yet currently underused, information from routine MRI acquisitions have been explored. The present study aimed to evaluate the accuracy of three automatic tumour segmentation programmes using the BRATS validation 2018 glioma dataset (n=66). The performance of BraTumIA was assessed across an adult glioma dataset from Imperial College Healthcare NHS Trust (n=13) and a paediatric brain tumour dataset from Great Ormond Street Hospital NHS Foundation Trust. DeepMedic segments adult gliomas more accurately than BraTumIA and ONCOhabitats. However, BraTumIA provides a more user-friendly segmentation tool.

This abstract and the presentation materials are available to members only; a login is required.

Join Here