Anatomy-guided PET reconstruction using l ₁ bowsher prior

Abstract

Advances in simultaneous positron emission tomography/magnetic resonance imaging (PET/MRI) technology have led to an active investigation of the anatomy-guided regularized PET image reconstruction algorithm based on MR images. Among the various priors proposed for anatomy-guided regularized PET image reconstruction, Bowsher’s method based on second-order smoothing priors sometimes suffers from over-smoothing of detailed structures. Therefore, in this study, we propose a Bowsher prior based on the l 1 -norm and an iteratively reweighting scheme to overcome the limitation of the original Bowsher method. In addition, we have derived a closed solution for iterative image reconstruction based on this non-smooth prior. A comparison study between the original l 2 and proposed l 1 Bowsher priors was conducted using computer simulation and real human data. In the simulation and real data application, small lesions with abnormal PET uptake were better detected by the proposed l 1 Bowsher prior methods than the original Bowsher prior. The original l 2 Bowsher leads to a decreased PET intensity in small lesions when there is no clear separation between the lesions and surrounding tissue in the anatomical prior. However, the proposed l 1 Bowsher prior methods showed better contrast between the tumors and surrounding tissues owing to the intrinsic edge-preserving property of the prior which is attributed to the sparseness induced by l 1 -norm, especially in the iterative reweighting scheme. Besides, the proposed methods demonstrated lower bias and less hyper-parameter dependency on PET intensity estimation in the regions with matched anatomical boundaries in PET and MRI. Therefore, these methods will be useful for improving the PET image quality based on the anatomical side information.

Evaluation of MR anatomically-guided PET reconstruction using a convolutional neural network in PSMA patients

“…This sometimes may result in over-smoothing of lesions with detailed structure, and this can be observed from the SD in tumors of the 32 patients (figure 6(b)), where the SD median from BCNN and BCNN0 are lower than the SD median of the MRAC, indicating over-smoothing. Most recently, the over-smoothing was addressed using an iterative weighting scheme and a first-order smoothing l 1 -norm by Kang and Lee (2021), and they reported that the mean uptake level of VOI was preserved while better detectability of small lesions was achieved. It is also worth mentioning that some tumor VOIs could have homogeneous texture where a lower SD is expected, while the SD will be higher in VOIs of heterogenies tumors.…”

Simultaneous PET/MRI: The future gold standard for characterizing motor neuron disease—A clinico-radiological and neuroscientific perspective

“…• Novel regularized PET image reconstruction techniques based on anatomical priors derived from concurrent MRI significantly improve PET image quality (173)(174)(175).…”

Anatomy-guided multi-resolution image reconstruction in PET

“…In PET, a number of different priors have been developed (Ehrhardt et al 2019, Kang andLee 2021). These priors have been mainly applied to the pixel model, however, a Gibbs prior (Brankov et al 2004) has been applied to a triangular mesh.…”