HEAD AND NECK IMAGING METHOD AND DEVICE BASED ON DEEP PRIOR LEARNING

摘要

Disclosed are a head and neck imaging method and device based on deep prior learning. The method comprises the following steps: obtaining a head and neck magnetic resonance image to be reconstructed (101); inputting the image into a pre-established plural convolutional neural network provided with plural residual blocks (102); and reconstructing the image by means of the plural convolutional neural network to obtain an artifact-free high resolution head and neck image (103). The described solution solves the problem in the existing head and neck imaging of not being able to guarantee imaging precision and imaging time at the same time, thereby achieving the technical effect of effectively reducing the imaging time when guaranteeing the imaging precision.