Abstract #3856

The video domain transfer deep learning network with error correction for Dixon Imaging with consistent slice-to-slice water and fat separation

Jong Bum Son¹, David Rauch¹, Bikash Panthi¹, Zijian Zhou¹, Benjamin Musall¹, Marion Scoggins², Mark Pagel³, and Jingfei Ma¹

¹Imaging Physics Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, ²Diagnostic Radiology Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, ³Cancer Systems Imaging Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States

We applied a video processing deep-learning neural network (originally developed to synthesize high resolution photorealistic video from a time series of dancing poses, semantically segmented street-view labels, or human face outline sketches) to Dixon imaging to present combined benefits of 2D and 3D networks. The developed Dixon Video Domain Transfer Generative Adversarial Network (DixonVDTGAN) could create slice-to-slice consistent water images with reduced demand on GPU memory. It could also successfully correct deep-learning processing errors for robust water and fat signal separation under two assumptions that the deep-learning processing errors are localized and the image phase is spatially smooth.

How to access this content:

For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.

After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.

After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.

Click here for more information on becoming a member.