Unsupervised registration for liver CT-MR images based on the multiscale integrated spatial-weight module and dual similarity guidance.

PURPOSE: Multimodal registration is a key task in medical image analysis. Due to the large differences of multimodal images in intensity scale and texture pattern, it is a great challenge to design distinctive similarity metrics to guide deep learning-based multimodal image registration. Besides, since the limitation of the small receptive field, existing deep learning-based methods are mainly suitable for small deformation, but helpless for large deformation. To address the above issues, we present an unsupervised multimodal image registration method based on the multiscale integrated spatial-weight module and dual similarity guidance. METHODS: In this method, a U-shape network with our multiscale integrated spatial-weight module is embedded into a multi-resolution image registration architecture to achieve end-to-end large deformation registration, where the spatial-weight module can effectively highlight the regions with large deformation and aggregate discriminative features, and the multi-resolution architecture further helps to solve the optimization problem of the network in a coarse-to-fine pattern. Furthermore, we introduce a special loss function based on dual similarity, which represents both global gray-scale similarity and local feature similarity, to optimize the unsupervised multimodal registration network. RESULTS: We verified the effectiveness of the proposed method on liver CT-MR images. Experimental results indicate that the proposed method achieves the optimal DSC value and TRE value of 92.70 ± 1.75(%) and 6.52 ± 2.94(mm), compared with other state-of-the-art registration algorithms. CONCLUSION: The proposed method can accurately estimate the large deformation field by aggregating multiscale features, and achieve higher registration accuracy and fast registration speed. Comparative experiments also demonstrate the effectiveness and generalization ability of the algorithm.

[The preparation and identification of diagnostic recombinant glycerol kinase].

In order to establish an efficient and low-cost production procedure of recombinant glycerol kinase (r-GK), we expressed the r-GK gene at high level in E. coli by induction with lactose on a large-scale fermentation of 300L. The results showed that the biomass concentration reached OD600 of 42 and the expression of r-GK in E. coli accounted for about 30% of total soluble protein. The cell-free extract was processed by selective thermo-denaturation and then purified with Ni sepharose FF column chromatography. Finally, highly purified r-GK was obtained and its purity reached 97% by using analysis on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), polyacrylamide gel electrophoresis (PAGE) and gradient polyacrylamide gel electrophoresis (Gradient PAGE). Further identification study showed that the molecular weight of r-GK was 120kDa with two subunit of 58kDa. Contaminants of NADH oxidase and catalase were not detected in the sample pool of r-GK. The purified r-GK was able to retain about 85% of its initial activity at 4 degrees C for 30 days. After lyophilized, it can retain 93% of its initial activity at 4 degrees C for one year.