Abstract #4690

Utilising multi-parametric MRI to non-invasively predict tumour type in paediatric neuro-oncological disease: a multi-centre study.

James Timothy Grist¹, Stephanie Timothy Withey², Lesley MacPherson³, Adam Oates⁴, Stephen Timothy Powell², Jan Novak⁵, Laurence Abernethy⁶, Barry Pizer⁷, Ricahrd Grundy⁸, Simon Bailey⁹, Dipayan Mitra⁹, Theodoros N Arvantis¹⁰, Dorothee P. Auer⁸, and Andrew C Peet²

¹University of Birmingham, BIRMINGHAM, United Kingdom, ²University of Birmingham, Birmingham, United Kingdom, ³Birmingham Women's and CHildren's NHS foundation trust, Birmingham, United Kingdom, ⁴Birmingham Women's and Children's NHS foundation trust, Birmingham, United Kingdom, ⁵Aston University, Birmingham, United Kingdom, ⁶Alder Hey Children's NHS foundation trust, Liverpool, United Kingdom, ⁷Institute of Translation Medicine, University of Liverpool, Liverpool, United Kingdom, ⁸University of Nottingham, Nottingham, United Kingdom, ⁹Royal Victoria Infirmary, Newcastle, United Kingdom, ¹⁰University of Warwick, Warwick, United Kingdom

This study focuses on utilising supervised Machine Learning to combine both diffusion and perfusion weighted imaging to discriminate between the three most common pediatric brain tumour types: Pilocytic Astrocytoma, Ependymoma, and Medulloblastoma.

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

Join Here