1750

View Abstract / View Presentation

Multi-channel and multi-group-based CNN Image Reconstruction using Equal-sized Multi-resolution Image Decomposition

Shohei Ouchi¹ and Satoshi ITO¹
¹Utsunomiya University, Utsunomiya, Japan

The transformed image domain CNN-CS method were proposed based on the equal-sized multi-resolution image decomposition method. Reconstruction experiments showed this CNN-CS could reconstruct sharp images that have strong phase variation.

Fig.5 Comparison of reconstructed images between eFREBAS-CNN, Image Domain CNN, ADMM-CSNet and CS iterative. (h) – (m) are the enlarged images of (a) – (f), respectively.

Fig.3 Network architecture and reconstruction flow of eFREBAS-CNN. Three groups are reconstructed by three CNNs, respectively. U-Net2 and U-Net3 that treats Group B and Group C are input lower frequency sub-images as shown in Fig.2.