Computed tomography (CT) iteration reconstruction method for asymmetrical detector

Abstract

The invention relates to a computed tomography (CT) iteration reconstruction method for an asymmetrical detector. The CT iteration reconstruction method is characterized by comprising the following steps of: performing air correction, nonlinear correction, crosstalk correction, defocus correction and calibration correction on raw data acquired by a computed tomography scanner to obtain projection data; judging whether the projection data is CT scan data or helical scan data, and if the projection data is the CT scan data, grouping the CT scan data; calculating a projection weight to obtain a weight matrix; performing ordered subset expectation maximization (OSEM) iteration orthographic and back projection according to the weight matrix; judging whether iteration ending conditions are met, if so, ending the iteration orthographic and back projection to obtain an initial CT image; and performing postprocessing on the initial CT image to obtain the reconstructed CT image. By the method, the iteration back projection reconstruction is carried out by directly using incomplete projection data, artifact can be relived, noise level difference caused by data inconsistency due to interpolation is inhibited, and the CT iteration reconstruction method is mainly suitable for less-than-four-row CT detectors.

Description

technical field [0001] The invention relates to medical image processing technology, in particular to a CT iterative reconstruction method with an asymmetric detector. Background technique [0002] In CT scanning, in order to increase the scanning field of view or reduce the radiation dose, some detector modules are usually added or subtracted on the side of the central channel of the detector to form an asymmetric detector array, such as figure 1 As shown, some detectors are reduced; compared with the symmetric detector structure, the asymmetric structure in the figure is equivalent to the lack of some projection data, resulting in a reduced reconstruction field of view. Such as figure 2 As shown, the fan beam geometry is reorganized into a parallel beam geometry, where FOV_In is [-γ m +α,γ m -α] can cover the field of view, FOV_Out is the part of the field of view between the complete scanning field of view and FOV_In. This figure is only a schematic diagram. A channe...

AI Technical Summary

Problems solved by technology

Because it is based on statistical laws, the MLEM reconstruction method has a good anti-noise ability, especially when dealing with poor statistical data, it can show its superiority over the analytical method, but this method still has iterative methods Disadvantages such as large amount of computation and long computation time

Images