X-ray contrast image simulation method based on gpu parallel operation

Abstract

The invention provides a simulation method for an X ray angiography image based on GPU parallel computation. The method can improve the validity of high simulation results and time efficiency of simulation algorithms. The method includes constructing a rotary angiography imaging model; dividing a coronary artery of a computed tomography (CT) image by using an area increasing method and simulating a coronary artery reinforced three-dimensional image through weight optimization of a coronary artery CT value; building a perspective projection model according to parameter setting of a rotary angiography imaging system and calculating a mapping relation between three-dimensional image space and two-dimensional image space; constructing an energy attenuation model and calculating energy distribution of X rays; and initializing texture data of the GPU by using a voxel coordinate of a three-dimensional simulation image and calculating a simulation angiography image by using the perspective projection model and the energy attenuation model.

Description

technical field [0001] The invention relates to an X-ray contrast image simulation method based on GPU parallel computing, which is suitable for the evaluation of the advantages of reconstruction algorithms and the training of clinical medical personnel and other fields. Background technique [0002] Coronary angiography is known as the "gold standard" in the diagnostic and therapeutic imaging methods of cardiovascular diseases due to its high resolution and high contrast. Ways to obtain contrast images include single-plane, dual-plane and rotational contrast imaging methods. Rotational contrast imaging can achieve multi-angle irradiation, which reduces the probability of misdiagnosis of stenotic lesions, and facilitates doctors to understand the degree of vascular stenosis and formulate appropriate treatment plans. However, clinical coronary angiography images still generally have the following problems: (1) serious crossing and occlusion caused by the intricate structure ...

AI Technical Summary

Problems solved by technology

However, clinical coronary angiography images still generally have the following problems: (1) serious crossing and occlusion caused by the intricate structure of blood vessels; (2) a large amount of spatial information is lost in two-dimensional angiography images; (3) there is a large amount of background noise in angiography images

[0007] 1. It is not possible to perform all-round imaging on contrast images to achieve a comprehensive diagnosis of patients;

[0008] 2. Compared with the real contrast images, the simulation results are less realistic;

[0009] 3. The time efficiency of simulated contrast images is low

Images