Learning Multi-Modal Volumetric Prostate Registration with Weak Inter-Subject Spatial Correspondence
This work provides a more accessible and less error-prone method for medical image registration, particularly for prostate interventions, by reducing the reliance on tedious voxel-to-voxel ground truth data.
This paper addresses the challenge of multi-modal volumetric prostate registration using weakly-supervised learning, optimizing a convolutional neural network with segmentation masks instead of explicit deformation fields. By combining multiscale Dice similarity coefficient and signed distance map similarities, and introducing an auxiliary input for prostate location, the method achieves registration quality comparable to state-of-the-art deep learning approaches.
Recent studies demonstrated the eligibility of convolutional neural networks (CNNs) for solving the image registration problem. CNNs enable faster transformation estimation and greater generalization capability needed for better support during medical interventions. Conventional fully-supervised training requires a lot of high-quality ground truth data such as voxel-to-voxel transformations, which typically are attained in a too tedious and error-prone manner. In our work, we use weakly-supervised learning, which optimizes the model indirectly only via segmentation masks that are a more accessible ground truth than the deformation fields. Concerning the weak supervision, we investigate two segmentation similarity measures: multiscale Dice similarity coefficient (mDSC) and the similarity between segmentation-derived signed distance maps (SDMs). We show that the combination of mDSC and SDM similarity measures results in a more accurate and natural transformation pattern together with a stronger gradient coverage. Furthermore, we introduce an auxiliary input to the neural network for the prior information about the prostate location in the MR sequence, which mostly is available preoperatively. This approach significantly outperforms the standard two-input models. With weakly labelled MR-TRUS prostate data, we showed registration quality comparable to the state-of-the-art deep learning-based method.