MRI Image-to-Image Translation for Cross-Modality Image Registration and Segmentation

We develop a novel cross-modality generation framework that learns to generate predicted modalities from given modalities in MR images without real acquisition. Our proposed method performs image-to-image translation by means of a deep learning model that leverages conditional generative adversarial networks (cGANs). Our framework jointly exploits the low-level features (pixel-wise information) and high-level representations (e.g. brain tumors, brain structure like gray matter, etc.) between cross modalities which are important for resolving the challenging complexity in brain structures. Based on our proposed framework, we first propose a method for cross-modality registration by fusing the deformation fields to adopt the cross-modality information from predicted modalities. Second, we propose an approach for MRI segmentation, translated multichannel segmentation (TMS), where given modalities, along with predicted modalities, are segmented by fully convolutional networks (FCN) in a multi-channel manner. Both these two methods successfully adopt the cross-modality information to improve the performance without adding any extra data. Experiments demonstrate that our proposed framework advances the state-of-the-art on five MRI datasets. We also observe encouraging results in cross-modality registration and segmentation on some widely adopted datasets. Overall, our work can serve as an auxiliary method in clinical diagnosis and be applied to various tasks in medical fields.