A method for image reconstruction and system error correction for cone CT

By using pipelined collaborative operations of CUDA and multithreaded CPUs during the CT reconstruction process, and calculating hardware error parameters through the two-bit grayscale entropy of the tomographic diagram, the problems of slow reconstruction speed and high error correction cost are solved, and more efficient computing resource usage and lower manufacturing costs are achieved.

CN114283220BActive Publication Date: 2025-03-28WUHAN GREENPHENO SCI & TECH CO LTD
1 Cites 0 Cited by

Patent Information

Application Number
CN202111606871.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-20
Publication Date
2025-03-28
Estimated Expiration
2041-12-20

AI Technical Summary

Technical Problem

There are problems such as slow speed, large computing resource utilization and high system error correction cost during the cone CT reconstruction process. The existing technology is difficult to speed up the reconstruction speed and reduce performance requirements while maintaining image accuracy, and it is impossible to calculate system hardware errors through software algorithms.

Method used

The CT image is reconstructed by pipelined collaborative operation of CUDA and multithreaded CPU, and the hardware error parameters of the imaging system are calculated through the two-bit grayscale entropy of the tomographic map, and these parameters are corrected during backprojection reconstruction to obtain the final CT reconstruction image.

Benefits of technology

It improves the speed of CT reconstruction, reduces the use of computing resources, and calculates hardware errors through software algorithms, reduces manufacturing and maintenance costs, and increases the system's tolerance for hardware errors.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present invention discloses a method for image reconstruction and system error correction for conical CT. The steps of the method are: firstly, perform CT scanning on the sample to obtain raw data and perform bright / dark current correction at the same time; based on the corrected data, use CUDA and multi-threaded CPU to quickly obtain the reconstructed CT raw image; and through segmented reconstruction and streaming transmission, reduce the requirements for memory and video memory while reducing the total time; calculate the ray source optical axis offset value, detector center deviation value, object distance and detector distance of the imaging system according to the two-bit grayscale entropy of the tomographic image at a specific position; correct the above parameters during back-projection reconstruction and obtain the final CT reconstructed image. Compared with the prior art, the present invention improves the speed and quality while also reducing the requirements for the operation space and the accuracy requirements of each component of the CT imaging system, and completes the calculation of spatial parameters without using calibration objects.
Need to check novelty before this filing date? Find Prior Art