Impact of Spherical Coordinates Transformation Pre-processing in Deep Convolution Neural Networks for Brain Tumor Segmentation and Survival Prediction

Pre-processing and Data Augmentation play an important role in Deep Convolutional Neural Networks (DCNN). Whereby several methods aim for standardization augmentation of the dataset, we here propose a novel method aimed to feed DCNN with spherical space transformed input data that could better facilitate feature learning compared standard Cartesian images volumes. In this work, coordinates transformation has been applied as preprocessing that, used conjunction normal MRI volumes, improves accuracy brain tumor segmentation patient overall survival (OS) prediction on Brain Tumor Segmentation (BraTS) Challenge 2020 dataset. The LesionEncoder framework then automatically extract features from models, achieving 0.586 OS validation set, which is one best results according BraTS leaderboard.