Multi-modal MRI synthesis method for representing GAN based on potential information

Abstract

The invention discloses a multi-modal MRI synthesis method based on potential information representation GAN. The multi-modal MRI synthesis method comprises the following steps: inputting collective information of different MRI modals into a generation network; the generation network respectively extracting potential information representation of multiple MRI modes by using different encoders; further transmitting the extracted potential information representation to a potential space processing network for integrated processing; obtaining a corresponding synthetic target mode by a decoder toserve as a synthetic image; jointly inputting the synthesized image and the real image into an identification network; distinguishing the real image from the generated image by the authentication network. According to the invention, a plurality of input modes can be flexibly received and synthesized, so that information loss can be effectively avoided, the fidelity of a synthesized image is effectively improved, and a high-quality image which truly reflects a detection part is obtained. The method is wide in application range, high in calculation efficiency and good in actual application effect.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a multimodal MRI synthesis method based on latent information characterizing GAN. Background technique [0002] Magnetic resonance imaging (MRI), as a noninvasive imaging technique, has become the main imaging modality for studying neuroanatomy. Different modalities of the same anatomy can be created by using different pulse sequences and parameters to generate different tissue contrast images. However, in practice, due to time constraints, some sequences are lost; some modalities suffer from random noise and unintended image artifacts, resulting in poor image quality. The above factors make the number of comparison images available for the same detected person always limited. Therefore, the missing or damaged modality needs to be synthesized with other modalities that have been successfully obtained; the synthesized modality can not only replace the missin...

AI Technical Summary

Problems solved by technology

Since different modes exhibit different physical states, these methods greatly reduce the efficiency of extracting information from each mode, resulting in unsatisfactory synthesis results.

Images