PET (Polyethylene terephthalate) three-dimensional image reconstruction method and device

Abstract

The invention provides a PET (Polyethylene terephthalate) three-dimensional image reconstruction method and device. The method comprises the following steps: initializing a three-dimensional image to cause the pixel value of a pixel, which is positioned on a three-dimensional matrix in a first coordinate system (X, Y, Z), of the initialized three-dimensional image to be f1(x,y,z); selecting one line-of-response subset from a plurality of line-of-response subsets obtained from a detector; successively selecting line-of-response groups from the selected line-of-response subset, and independently carrying out preset operation to each selected line-of-response group, wherein the preset operation comprises the step of obtaining the pixel value g1(s,t,z) of each pixel of the first pixel group which is positioned in a second coordinate system (S,T,Z) and corresponds to the selected according to the pixel value f1(x,y,z) of the pixel in the first coordinate system (X, Y, Z), and the pixels of the first pixel group are positioned on a plurality of lines of response corresponding to the selected line-of-response group. Three-dimensional image reconstruction efficiency can be improved.

Description

technical field [0001] The invention relates to the technical field of medical image processing, in particular to a PET three-dimensional image reconstruction method and device. Background technique [0002] Ordered Subset Expectation Maximization (OSEM) is currently the main iterative algorithm for 3D image reconstruction in PET clinics. This algorithm divides the projection data into n subsets, and only one subset is used for each reconstruction. Correction is performed, and the reconstructed image is updated once, and all subsets correct the projection data once, which is called an iteration; when the stop rule is met, the iteration is terminated, otherwise the new reconstructed image is used as the initial image to enter the next iteration. [0003] However, the defect of the OSEM algorithm is that the calculation amount is large and the calculation speed is slow. Specifically, each time the reconstructed image is updated, the correction coefficient of each response lin...

AI Technical Summary

Problems solved by technology

When the reconstructed image is small and the number of response lines in the system is small (such as two-dimensional reconstruction), the correction coefficients of all response lines to all pixels can be pre-calculated, and then called during reconstruction by look-up table method; but in high-resolution When reconstructing at a high rate (such as 3D reconstruction), the reconstructed image is very large and there are a large number of response lines in the system. The existing data storage media cannot meet the requirements, and the correction coefficient of each response line to each pixel can only be calculated in real time during reconstruction.

Images