Depth self-supervision fast magnetic resonance imaging method and device based on cross-modal prior

Abstract

The invention discloses a depth self-supervision fast magnetic resonance imaging method and device based on cross-modal prior, and particularly relates to the technical field of medical imaging, the method comprises the steps that first image data and second image data are acquired, and the first image and the second image belong to images of different modalities; the first image data and the second image data are sent to a cross-modal fusion network formed by cascading at least two cross fusion modules for cross-modal fusion, and cross-modal fusion features are generated; and performing image reconstruction based on the cross-modal fusion features, and outputting a reconstructed image. Two different modes of image data are sent into a cross-modal fusion network for cross-modal cross fusion, and first image data and second image data are subjected to data cross fusion by means of the cross-modal cross fusion network, so that the purpose of increasing training data is achieved. The problem that a large amount of training data needs to be prepared in advance for image reconstruction in the prior art is relieved, and the duration of magnetic resonance imaging is shortened.

Images