Brain Tumor Segmentation with Uncertainty Estimation and Overall Survival Prediction

Accurate segmentation of different sub-regions of gliomas including peritumoral edema, necrotic core, enhancing and non-enhancing tumor core from multimodal MRI scans has important clinical relevance in diagnosis, prognosis and treatment of brain tumors. However, due to the highly heterogeneous appearance and shape, segmentation of the sub-regions is very challenging. Recent development using deep learning models has proved its effectiveness in the past several brain segmentation challenges as well as other semantic and medical image segmentation problems. Most models in brain tumor segmentation use a 2D/3D patch to predict the class label for the center voxel and variant patch sizes and scales are used to improve the model performance. However, it has low computation efficiency and also has limited receptive field. U-Net is a widely used network structure for end-to-end segmentation and can be used on the entire image or extracted patches to provide classification labels over the entire input voxels so that it is more efficient and expect to yield better performance with larger input size. In this paper we developed a deep-learning-based segmentation method using an ensemble of 3D U-Nets with different hyper-parameters. Furthermore, we estimated the uncertainty of the segmentation from the probabilistic outputs of each network and studied the correlation between the uncertainty and the performances. Preliminary results showed effectiveness of the segmentation model. Finally, we developed a linear model for survival prediction using extracted imaging and non-imaging features, which, despite the simplicity, can effectively reduce overfitting and regression errors.