Peer-Reviewed Publication
J Neurotrauma2026;43(15-16):1238-1249.August 1, 2026Journal Article
Optimization and External Validation of a Deep Learning Model for Segmentation and Quantification of Traumatic Brain Injury Lesions on Head Computed Tomography.
François Mathieu1,2, Carolina Piçarra3, Armaan Malhotra1,4,5, Julia Srbely4, Miguel Monteiro3, Stefan Winzeck3, Christopher Smith4, Virginia Newcombe6, David Menon6, Christopher Witiw1,4,5, Damon Scales2,5,7, Farhad Pirouzmand1,7, Ben Glocker3
1Department of Surgery, Division of Neurosurgery, University of Toronto, Toronto, Ontario, Canada.
2Interdepartmental Division of Critical Care Medicine, University of Toronto, Toronto, Ontario, Canada.
3Department of Computing, Biomedical Image Analysis Group, Imperial College London, London, UK.
4Li Ka Shing Knowledge Institute, St Michael's Hospital, Toronto, Ontario, Canada.
5Institute of Health Policy, Management and Evaluation, Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada.
6Department of Medicine, University Division of Anaesthesia and PACE Section, University of Cambridge, Cambridge, UK.
7Department of Critical Care Medicine, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada.
Abstract
Current methods for radiological classification in traumatic brain injury (TBI), such as the Marshall and Rotterdam score, provide an incomplete description of intracranial lesion burden, rely on time-consuming manual assessments by experts, and are prone to intra- and interrater disagreement. To circumvent these limitations, we previously proposed the Brain Lesion Analysis and Segmentation Tool f…
Create a free account to keep reading
Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.
Want unlimited research access? See Pro plans