作者: M Defrise , S Kuijk , F Deconinck
DOI: 10.1088/0031-9155/33/1/004
关键词: Computer vision 、 Coincidence 、 Reconstruction algorithm 、 Plane (geometry) 、 Space (mathematics) 、 Positron 、 Physics 、 Frequency domain 、 Detector 、 Iterative reconstruction 、 Artificial intelligence
摘要: The authors propose a new truly three-dimensional reconstruction algorithm, which is based on the Fourier space filtering of backprojected image, but overcomes requirement invariance, usually forces one to reject significant fraction measured projections. algorithm has been applied images from list mode coincidence data recorded by positron camera pair rotating multi-wire chambers. Results obtained with simulated demonstrate improvement reconstructed due better utilisation available data.