China
Africa
Explore chapters and articles related to this topic
Principles and instrumentation of SPECT/CT
Published in Yi-Hwa Liu, Albert J. Sinusas, Hybrid Imaging in Cardiovascular Medicine, 2017
The FBP algorithm inherently assumes that the projection data are consistent. In other words, FBP assumes that projection at each different angle is a sum of the same object. However, this assumption is violated by noise. Each projection is a sum of the object plus noise and the noise is random and hence different in each projection. Other factors like attenuation, scatter, distance-dependent collimator resolution, and patient motion (Section 1.5) introduce further inconsistencies between the projections. This has led to the development of alternative methods of reconstruction. Because the data are inconsistent, the solution is often pursued using iterative methods (Figure 1.7) that try to find the solution that is the best compromise between the measured data, something that minimizes the errors. The most popular algorithm used clinically is the ordered subset expectation maximization (OSEM) algorithm (Hudson and Larkin 1994) that is a variant on the maximum likelihood expectation maximization algorithm that was introduced into nuclear medicine by Shepp and Vardi (1982). This algorithm is built on the assumption that the noise in the data is Poisson distributed. The idea of iterative reconstruction is illustrated in Figure 1.7. An estimate is made of the distribution of activity. Having no other information available, it is usually initially assumed that the activity is uniformly distributed throughout the FOV, though any positive, nonzero distribution is a valid starting point. Using an understanding of the geometry of the camera and how photons travel from points within the FOV to the detector, a set of projections are computed that show what the camera would have measured given the current estimate of the activity distribution (the forward projection). Then the calculated projections are compared to the ones measured with the camera. Where the calculated values are higher than the measured ones, we reduce the activity in the object that contributes to that part of the projections. Where the calculated values are lower, we increase our estimate. This is done by backprojecting the ratio of measured over calculated projections. The backprojected ratios are used to rescale the original estimate and generate a better one. We then repeat or iterate the process until the estimate stops changing, giving us our reconstructed image.
A novel cardiac SPECT system and imaging method
Published in The Imaging Science Journal, 2020
Jinhua Sheng, Yangjie Ma, Rougang Zhou, Xun Li, Luyun Wang, Yuchen Shi
In summary, we propose a novel SPECT system with an elliptical orbit. Under the structure of the proposed model, the distance between the detector and the ROI becomes smaller, which can effectively improve the SNR. At the same time, the slit-slat collimator is used to collimate the emitted photons. With the geometry and slit-slat collimator used in this model, the efficiency and sensitivity of the system can be effectively improved. The geometric efficiency of the proposed SPECT is about twice that of the conventional SPECT system, but the scanning time is only about one tenth. Finally, the reconstruction process under the proposed SPECT model is simulated. The reconstruction algorithm is modified on the traditional OSEM reconstruction algorithm by adding attenuation correction, body support and TV minimization constraints. The results show that the lack of these constraints may lead to serious deviations in the reconstruction results. Simultaneous sampling coverage angle is also one of the important factors affecting the reconstruction results. According to the reconstruction results, the quality of reconstruction results is best when the sampling coverage angle is 210°. This results might provide a new effectively way for understanding of the metabolism of the human heart in clinical treatment.