Correcting method for geometric parameters of tomography

Abstract

The invention provides a correcting method for geometric parameters of tomography and relates to the technical field of tomography. The correcting method comprises the following steps: collecting all-angle projected images in the tomography; respectively performing negative logarithm operation on the collected all-angle projected images; superposing the all-angle projected images after being subjected to the negative logarithm operation, thereby obtaining a superposed image; utilizing the superposed image to establish a target function associated with the geometric parameters; and utilizing asimplex simulated annealing algorithm to perform global minimum optimization on the target function. According to the correcting method provided by the invention, the correcting for a plurality of geometric parameters is realized without using any precise mould; the superposed image is utilized, thereby being insensitive to noise and strong in noise resistance; and besides, the simplex simulated annealing algorithm is utilized in the method, so that the local minimal value is avoided and the precision and accuracy of simultaneously solving the plurality of geometric parameters are ensured.

Description

technical field [0001] The invention relates to the technical field of tomography, in particular to a method for calibrating geometric parameters of tomography, and is especially suitable for calibrating geometric parameters of a cone-beam X-ray tomography system. Background technique [0002] Cone beam computed tomography (CBCT) is a very important diagnostic and research tool in the fields of clinical medicine and biomedicine, including breast CT, dental CT, limb CT, small animal CT, etc. The geometric parameters of the CBCT system play a decisive role in obtaining reconstructed images with high resolution and low artifacts, and a slight deviation of the geometric parameters will lead to a serious decline in imaging quality. Therefore, the geometric parameters of the CBCT system need to be calibrated. [0003] The existing calibration methods of geometric parameters of CBCT system can be divided into two categories: [0004] One is off-line calibration, that is, before s...

AI Technical Summary

Problems solved by technology

This is because in the actual system, due to the influence of the environment, the mechanical structure will gradually deviate, resulting in dynamic changes in the geometric parameters of the system, especially in high-resolution systems, small changes in system geometric parameters will bring about a sharp change in performance In addition, some systems need to frequently adjust the field of view, magnification, etc., which makes the offline calibration method cumbersome and unreliable

Chinese invention patent ZL 200810136662.6 obtains the position of the rotation center by binarizing the projection image and then finding the center of gravity of the image; Thus the position of the center of rotation is obtained; the article Medical Physics 36 (2009) 48-58 uses the sinogram to solve the position of the rotation axis and the inclination angle in the plane; the article Physics In Medicine and Biology53 (2008) 3841-3861 uses the sinogram to obtain The cost function associated with the geometric parameters, and then use the simplex algorithm to obtain the position of the rotation axis and the two inclination angles of the detector, but this method is not suitable for systems with large cone angles, and the calculation results are sensitive to the selection of the region of interest. It is complicated, and the simplex algorithm used is easy to fall into the local minimum and thus cannot get the correct result; the article Medical Physics 38 (2011) 4934-4945 is to substitute the geometric parameters into the CT reconstruction algorithm, and compare the sharpness of the edge of the reconstructed image The position of the rotation axis, the two inclination angles of the detector, and the distance from the ray source to the detector are obtained through the degree of optimization. In the process of solving, this method adopts an exhaustive method without optimization, and every adjustment of the geometric parameters requires Reconstructing an image is a complicated and time-consuming process

The overall problem of the existing online calibration method is that the establishment of the objective function is very difficult, the solution is complicated, and the accuracy is slightly lower than that of offline calibration.

Images