x-ray contrast image simulation based parallel computing gpu

The present invention provides a X-ray contrast image simulation method of GPU-based parallel computing can be improved time efficiency and fidelity simulation results of simulation algorithm. 1 comprises a main building rotary contrast imaging model; 2, using the region growing method coronary CT images is divided by the weight of heavy coronary CT value optimization, simulated three-dimensional image of coronary enhanced; 3, according to the rotation angiography imaging system parameter settings established perspective projection model to calculate a three-dimensional image space to two-dimensional imaging space mapping relationship; 4, energy attenuation model constructed to calculate energy distribution of the transmitted X-rays; 5, voxel coordinates 3D simulation image of the GPU texture initialization data, calculating a contrast image simulation models using perspective projection and parallel energy attenuation model.