Research On Signal Readout And Evaluation Technique For Edge-on X-ray Detector Pixels

Medical computed tomography(CT)obtains a tomographic image of the tissue being scanned by scanning the animal body by radiography,which is widely used to prevent and treat diseases.Spectral CT can project the material through a circular scan by setting two or more effective energy spectrum intervals in advance.Compared with traditional CT,spectral CT can accomplish the separation and characterization of matter more accurately.X-ray detector and its signal readout techniques in CT imaging systems are key aspects of CT imaging systems.In order to meet the imaging requirement of dual-energy CT,the signal readout and imaging evaluation techniques of the Edge-on X-ray detector pixels were studied in this paper.Based on X-ray Spectrum Detecting and resolving methods,the readout system architecture and workflow of the output signal for detector pixels were first described in the article.The whole readout system includes four parts: a readout timing generation interface module,an AD control module,a reconfigurable accumulation module,and a universal serial bus(USB)communication interface control module.The function planning,interface design and module simulation of each module are described in detail.At the same time,this paper proposed a full-reference medical image quality assessment method based on visual perception for assessing the quality of the reconstructed dual-energy CT image quantitatively.The method used edge detection and Bayesian prediction model to divide the image into three regions with different perceptual features.By evaluating the three regions separately,the method is consistent with human subjective assessment.Finally,the design and implementation of key modules are carried out based on FPGA.The functional simulation verification of the readout system for X-ray detector,energy-resolving calculation,dual-energy CT image reconstruction and quantitative evaluation analysis of image quality were finished.The experimental results show that the signal readout system developed in this paper can realize the readout drive of the X-ray detector,the efficient transfer of photogenerated charges,readout of photogenerated charges,analog-to-digital conversion and group accumulation of signals,and output the data through the USB interface finally.The experimental results of the proposed medical image quality assessment method in the CSIQ database have a SROCC value of 0.8360,a KROCC value of 0.6269,a PLCC value of 0.8348,and an RMSE value of 0.1445.The experimental results show that the proposed method is superior to the commonly used medical image quality assessment method PSNR and the proposed method is similar to human subjective assessment.