Research On Quality Prediction And Parameter Optimization Of X-ray Imaging

X-ray imaging technique has been widely used in medical diagnosis,non-destructive examination,security monitoring and so on.Therefore,it is significant to study how to improve the imaging quality of X-ray equipment.Enhancing the imaging quality mainly depends on the improvement of hardware,however,when the hardware conditions are detennined it is also important to adjust the controllable parameters reasonably so as to achieve optimal imaging quality.This paper focuses on how to optimize and adjust the physical parameters of X-ray imaging equipment to improve the imaging quality and puts forward a new way to enhance the image quality.First of all,this paper puts forward a general idea on optimizing and setting the imaging parameters to improve the imaging quality of X-ray equipment by consulting relevant literature,an experimental platform is built and the X-ray detection samples with different thickness and density are prepared through lots of experiments to obtain sample data,which creates the conditions for subsequent analysis,model prediction and optimization work.Secondly,because the influencing factors of X-ray imaging are numerous and complex,such as anode voltage,X-ray source focal spot size,the density of sample and so on,the relevant influencing factors of X-ray imaging quality are studied in this paper.Using AHP method,this paper obtains the weights of various factors influencing on the imaging quality by programming calculation and puts them in order to extract the main factors of X-ray imaging quality,which lays the foundation for the establishment of the prediction model and the training of the parameters.Because the relationship between the X-ray imaging quality and its influencing factors is obviously nonlinear,the traditional mathematical method is not good for the prediction of X-ray imaging quality.Therefore,this paper sets up the three-layer BP neural network reflecting the relationship between X-ray imaging indexes and main factors.The sample data is used for learning and training the network based on PSO to obtain the prediction model of X-ray imaging quality.The model can be used to predict the imaging results affected by multi parameter combination.The predicted results obtained by the model have little difference with the actual results by experimental comparison.The average prediction error of the imaging sensitivity is about 2%and the average prediction error of the imaging resolution is about 1.8%.Finally,in order to optimize and adjust the physical parameters of X-ray imaging equipment influencing on the X-ray imaging quality,this paper uses optimal imaging quality as a target and working conditions of X-ray imaging device as a constraint to establish an optimization mathematical model of X-ray imaging based on the prediction model established before.And then this paper uses genetic algorithm to solve the model in order to realize the optimization of X-ray imaging parameters.The experimental results show that the imaging results obtained by genetic algorithms and the imaging results based on the artificial optimization method are basically the same,but the artificial optimization method to achieve the results needs to be constantly adjusted,it takes at least half a day.And the optimization method based on genetic algorithm needs just a few minutes.