An Optimized Registration Method Based on Distribution Similarity and DVF Smoothness for 3D PET and CT Images
Hongjian Kang,Huiyan Jiang,Xiangrong Zhou,Hengjian Yu,Takeshi Hara,Hiroshi Fujita,Yu-Dong Yao
DOI: https://doi.org/10.1109/ACCESS.2019.2961268
IF: 3.9
2020-01-01
IEEE Access
Abstract:A fusion image combining both anatomical and functional information obtained by registering medical images of two different modalities, Positron Emission Tomography (PET) and Computed Tomography (CT), is of great significance for medical image analysis and diagnosis. Medical image registration relies on similarity measure which is low between PET/CT image voxels and therefore PET/CT registration is a challenging task. To address this issue, this paper presents an unsupervised end-to-end method, DenseRegNet, for deformable 3D PET/CT image registration. The method consists of two stages: (1) predicting 3D displacement vector field (DVF); and (2) registering 3D image. In the 3D DVF prediction stage, a two-level similarity measure together with a deformation regularization is proposed as loss function to optimize network training.In the image registration stage, a resampler and a spatial transformer are utilized to obtain the registration results. In this paper, 663 pairs of Uptake Value (SUV) and Hounsfield Unit (Hu) patches of 106 patients, 227 pairs of SUV and Hu patches of 35 patients and 259 pairs of SUV and Hu patches of 35 patients are randomly selected as training, validation and test set, respectively. Normalized cross correlation (NCC), intersection over union (IoU) of liver bounding box and euclidean distance (ED) on landmark points are used to evaluate the registration results. Experiment results show that the proposed method, DenseRegNet, achieves the best results in terms of liver bounding box IoU and ED, and the second highest value of NCC. For a trained model, given a new pair of PET/CT images, the registration result can be obtained with only one forward calculation within 10 seconds. Through qualitative and quantitative analyses, we demonstrate that, compared with other deep learning registration models, the proposed DenseRegNet achieves improved results in the challenging deformable PET/CT registration task.