Deep learning segmentation of pediatric brain tumors using diffusion-weighted MRI: towards an early, in-vivo classification pipeline

Abstract

Diffusion-weighted MRI (DWI) can offer vital quantitative biomarkers to understand pediatric brain tumors to benefit clinical management. However, extraction of these markers requires delineation of the tumor margins, which is time-consuming, requires expertise and can still be highly variable. Automated approaches using deep learning to generate these tumor regions exist but, to our knowledge, few exist using only DWI as an input modality, with none in pediatrics. This retrospective study develops an automated segmentation approach, leveraging transfer-learning and multi-modal ensembling to tackle the issues specific to this DWI-only approach. Using the Imaging of Tumors study data we analysed data from 107 pediatric brain tumour patients. Using a 3D convolutional neural network (3D CNN), namely DeepMedic, to perform automatic segmentations, we demonstrate the benefit of these approaches for this task. We assess the accuracy of these predicted segmentations in comparison to the “ground truth” manual annotations, with a median Dice score of 0.63 achieved by the best performing model. The current study highlights the potential of this approach, to be implemented in future clinical decision support tools using DWI, but more work is needed to improve segmentation accuracy or establish current performance as ‘sufficient’ for the purposes that these segmentations are required.