Model calculations for PET sinograms
Linear and multilinear models can be applied to dynamic sinograms, producing parametric sinograms, which can be reconstructed to parametric images; either dynamic or parametric sinogram can be normalized and corrected attenuation (Maguire et al., 1997). Alternatively, the model fitting and iterative reconstruction could be combined (Matthews et al., 1997; Meikle et al., 1998; Tsoumpas et al., 2008a; Tsoumpas et al., 2008b; Wang & Qi, 2009; Angelis et al., 2011; Wang & Qi, 2012; Rahmim et al., 2012; Su et al., 2013; Wang & Qi, 2013; Rakvongthai et al., 2013; Yang et al., 2016), and even directly from the list mode data (Yan et al., 2012; Zhu et al., 2014). These can be of markedly better quality than parametric images that are calculated conventionally from reconstructed dynamic images (Kamasak et al., 2005 and 2014).
Bentourkia (2003) applied basis functions/spectral analysis method to compute the compartment model parameters of [11C]acetate for each sinogram bin, and reconstructed these to K1 (perfusion) and k2 (oxygen consumption) images. Gjedde-Patlak plot has been applied to sinogram data in several publications.
However, this procedure is not possible without thorough understanding of the sinogram file format and all the corrections applied to it. There is no standardization on how PET scanners work with the raw data, and users are not normally allowed to extract sinogram data out of the scanner computing unit, process it, and then import it back for reconstruction. Therefore, there has not been any breakthrough in these techniques, despite of their advantages.
Currently only few (and outdated) applications exist in TPC:
- Logan plot to produce DV or DVR sinogram (sensitive to noise, smoothing recommended)
- Sum sinogram frames, reconstruct an image, and use it to calculate parametric blood flow image using in vivo autoradiography methods
- Use the PatPar program (applying Gjedde-Patlak plot) on SUN workstations to calculate parametric sinogram, which can be reconstructed to a Ki image; current version requires that a 3-exp function is fitted to the plasma curve; method was used for example in the FDG brain study by Hakala et al. (2002).
Angelis GI, Tziortzi AC, Tsoumpas C. Reconstruction of linear kinetic parameters directly from projection PET data. J Phys Conf Series 2011; 317: 012002
Kamasak ME, Bouman CA, Morris ED, Sauer K. Direct reconstruction of kinetic parameter images from dynamic PET data. IEEE Trans Med Imaging 2005; 24(5): 636-650.
Kamasak ME, Christian BT, Bouman CA, Morris ED. Quality and precision of parametric images created from PET sinogram data by direct reconstruction: proof of concept. IEEE Trans Med Imaging 2014; 33(3): 695-707.
Maguire RP, Calonder C, Leenders KL. An investigation of multiple time/graphical analysis applied to projection data: theory and validation. J Comput Assist Tomogr. 1997; 21(2): 327-331.
Maguire RP, Spyrou NM, Leenders KL. Variance in parametric images: direct estimation from parametric projections. Phys Med Biol. 2000; 45: 91-102.
Morris ED, Kamasak ME, Christian BT, Cheng TE, Bouman CA. Visualizing all the fits: Evaluating the quality and precision of parametric images created from direct reconstruction of PET sinogram data. Proc Int Symp Biomed Imag. 2006; 291-294.
Rahmim A, Zhou Y, Tang J, Lu L, Sossi V, Wong DF. Direct 4D parametric imaging for linearized models of reversibly binding PET tracers using generalized AB-EM reconstruction. Phys Med Biol. 2012; 57: 733-755.
Rakvongthai Y, Ouyang J, Guerin B, Li Q, Alpert NM, El Fakhri G. Direct reconstruction of cardiac PET kinetic parametric images using a preconditioned conjugate gradient approach. Med Phys. 2013; 40(10): 102501.
Su K-H, Yen T-C, Fang Y-HD. A novel approach for direct reconstruction of parametric images for myocardial blood flow from PET imaging. Med Phys. 2013; 40: 102505.
Wang G, Qi J. Direct estimation of kinetic parametric images for dynamic PET. Theranostics 2013; 3(10): 802-815.
Updated at: 2017-11-22
Created at: 2003-12-11
Written by: Vesa Oikonen