CT image reconstruction method under condition that CT projection paths are reduced

Abstract

The invention relates to a CT image reconstruction method under the condition that CT projection paths are reduced. The CT image reconstruction method comprises the following steps: obtaining a seriesof continuous incomplete projections through scanning in different incomplete projection modes, and obtaining a CT image sequence through reconstruction; performing CT image spatial information datarecovery and image reconstruction by using a deep convolutional neural network; in a network training process, firstly, utilizing a data set of incomplete projection for pre-training a network, then,utilizing a complete data set for carrying out large-scale iterative training on the network, and obtaining an incomplete projection CT spatial information data recovery model; and inputting the realincomplete projection CT image slice into a data recovery model to obtain a high-quality CT reconstructed image. The CT image reconstruction method is reasonable in design, can obtain the high-qualityCT reconstructed image with high contrast, high definition and good detail description, and can greatly improve the reconstruction quality of the incomplete projection CT image under the condition that the X-ray tube current is reduced and the projection path is reduced.

Description

technical field [0001] The invention belongs to the technical field of computer images, in particular to a CT image reconstruction method under the condition of reduced CT projection paths. Background technique [0002] X-ray computed tomography (Computed Tomography, CT) technology has been widely used in industrial detection and medical diagnosis and other fields. However, since the high radiation dose of X-rays in CT examination will have adverse effects on human health, the development of low-dose CT is the focus and hotspot of research in this field. [0003] At present, one of the most direct ways to reduce the radiation dose of CT is to reduce the tube current or tube voltage of the X-ray tube, but directly reducing the tube current will increase the noise level of the projection data, which will lead to obvious noise plaques in the reconstructed image; Reducing the voltage of the X-ray tube will reduce the penetration ability of the X-ray, resulting in severe beam ha...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the deficiencies of the prior art, and propose a CT image reconstruction method under the condition of reduced CT projection path, which uses deep convolutional neural network to extract the spatial structure information of CT image sequence to realize CT image data of incomplete projection Restoration and reconstruction, solving the problem of insufficient amount of data required for incomplete projection reconstruction

Images