MICDIR: Multi-scale Inverse-consistent Deformable Image Registration using UNetMSS with Self-Constructing Graph Latent

Abstract

Image registration is the process of bringing different images into a common coordinate system -a technique widely used in various applications of computer vision, such as remote sensing, image retrieval, and most commonly in medical imaging. Deep Learning based techniques have been applied successfully to tackle various complex medical image processing problems, including medical image registration. Over the years, several image registration techniques have been proposed using deep learning. Deformable image registration techniques such as Voxelmorph have been successful in capturing finer changes and providing smoother deformations. However, Voxelmorph, as well as ICNet and FIRE, do not explicitly encode global dependencies (i.e. the overall anatomical view of the supplied image) and therefore can not track large deformations. In order to tackle the aforementioned problems, this paper extends the Voxelmorph approach in three different ways. To improve the performance in case of small as well as large deformations, supervision of the model at different resolutions have been integrated using a multi-scale UNet. To support the network to learn and encode the minute structural co-relations of the given image-pairs, a self-constructing graph network (SCGNet) has been used as the latent of the multi-scale UNet -which can improve the learning process of the model and help the model to generalise better. And finally, to make the deformations inverse-consistent, cycle consistency loss has been employed. On the task of registration of brain MRIs, the proposed method achieved significant improvements over ANTs and VoxelMorph, obtaining a Dice score of 0.8013±0.0243 for intramodal and 0.6211±0.0309 for intermodal, while VoxelMorph achieved 0.7747±0.0260 and 0.6071±0.0510, respectively.

Liver Segmentation using Turbolift Learning for CT and Cone-beam C-arm Perfusion Imaging

“…Moreover, to visualise and quantitatively evaluate the perfusion maps estimated based on the TST coefficients, the liver has to be segmented from the reconstructed TST coefficients and C-arm CBCT volumes. This warrants 3D-3D volumetric registration also with CT, which could also benefit from the liver segmentation (Avants et al, 2009;Chatterjee et al, 2022a). The potential usability of liver segmentation is summarised in Fig 1 . What makes the automatic liver segmentation a challenging task is the similar attenuation values of the liver and other surrounding organs, which consequently affects the estimation of liver contours (Lu et al, 2018).…”