Path Integral Method for X-ray Monte Carlo Simulation

Abstract

The invention discloses a path integral method for X-ray Monte Carlo simulation. The path integral method for the X-ray Monte Carlo simulation comprises the following steps: step a, a low-difference sequence with the length of m in a [0,1] 3n dimension is generated, wherein the n is the scattering times and the m is the number of samples; step b, scattering point sequences geometrically matched with a motif are generated through an isomorphic mapping of the low-difference sequence; step c, m photon paths are formed by the scattering point sequences on each detector grid point; step d, the probability density of passing through the photon paths to reach the detector grid points is determined; step e, the probability density is weighted and summed to obtain the scattering intensity of each detector grid point. According to the path integral method for the X-ray Monte Carlo simulation, errors caused by discontinuity on a boundary are eliminated, variances of the samples are lowered, calculation is finished with fewer samples under given accuracy, and the calculated amount is greatly reduced.

Description

technical field [0001] The invention relates to medical image processing technology, in particular to a path integration method for X-ray Monte Carlo simulation. Background technique [0002] Scattering numerical simulation of simple geometric models is the basis of X-ray (X-ray) scattering correction technology. Numerical simulation is an indispensable means to test the pros and cons of scattering correction technology. High-efficiency numerical simulation algorithm can not only speed up scattering The research and development of correction technology itself can also be directly developed into a kind of scattering correction technology. [0003] In the existing Monte Carlo X-ray scattering simulation, people use a series of pseudo-random numbers or quasi-random sequences to determine the exit fan angle, opening angle and penetration depth of photons, and track the probability of each beam of photons arriving at the detector grid. And add points. CRFD method as described i...

AI Technical Summary

Problems solved by technology

[0004] In this method, the randomly generated exit angles and penetration depths cannot ensure that the scattering points of each random photon fall inside the phantom, and these invalid scattering points cause a certain amount of additional calculation overhead, and the integrand at the phantom boundary does not Continuity leads to certain errors and cannot take full advantage of the low variance nature of quasi-random sequences

These errors need to provide a larger number of integration points to compensate, resulting in an increase in the amount of calculation

Images