Development and evaluation of a new high-TOF-resolution all-digital brain PET system

Fang, Lei; Zhang, Bo; Li, Bingxuan; Zhang, Xiangsong; Zhou, Xiaoyun; Yang, Jigang; Li, Ang; Shi, Xinchong; Liu, Yuqing; Kreissl, Michael; D’Ascenzo, Nicola; Xiao, Peng; Xie, Qingguo

doi:10.1088/1361-6560/ad164d

Cited by 3 publications

(3 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evaluate the effectiveness of DP-RI-Tikhonov in a pseudo-real scenario, we collected four high-quality [ 18 F]FDG brain PET images from a high resolution all-digital brain PET system (Fang et al 2024) as simulation phantoms, as shown in figures 1(f) and 7(f). For simplicity and ease of reading, a simulation phantom is used to analyze the effectiveness of DP-RI-Tikhonov in detail and the remaining three simulation phantoms are used to evaluate the generalization ability of DP-RI-Tikhonov.…”

Section: Application To Simulation Data (1) Experimental Settingmentioning

confidence: 99%

“…(2) Evaluation metrics: Due to the lack of the ground truth PET image x, contrast-to-noise ratio (CNR), CNR = (µ s − µ b ) /σ b , is adopted as the evaluation metric, where µ s and µ b are the mean of signal region and background region respectively, σ b is the standard deviation of background region. In general, the signal region and background region are manually drawn in a region with high activity distribution and a region with relatively uniform activity distribution, respectively (Jomaa et al 2018, Li et al 2023, Fang et al 2024. For better analysis, we calculate the CNR of the brain and the body separately.…”

Section: Application To Real Data (1) Experimental Settingmentioning

confidence: 99%

“…For better analysis, we calculate the CNR of the brain and the body separately. As shown in figures 8(a) and 9(a), similar to Fang et al (2024), we choose the subthalamic nucleus as the signal region and draw a sphere with a diameter of 10 mm on the white matter as the background region to calculate the CNR of the brain. As for the body, similar to Jomaa et al (2018), Li et al (2023), we choose the myocardial tissue as the signal region and draw two spheres with a diameter of 18 mm on the liver as the background regions.…”

Section: Application To Real Data (1) Experimental Settingmentioning

confidence: 99%

See 2 more Smart Citations

Deep denoiser prior driven relaxed iterated Tikhonov method for low-count PET image restoration

Chang,

D’Ascenzo,

Antonecchia

et al. 2024

Phys. Med. Biol.

View full text Add to dashboard Cite

Objective. Low-count positron emission tomography (PET) imaging is an efficient way to promote more widespread use of PET because of its short scan time and low injected activity. However, this often leads to low-quality PET images with clinical image reconstruction, due to high noise and blurring effects. Existing PET image restoration (IR) methods hinder their own restoration performance due to the semi-convergence property and the lack of suitable denoiser prior. Approach. To overcome these limitations, we propose a novel deep plug-and-play IR method called Deep denoiser Prior driven Relaxed Iterated Tikhonov method (DP-RI-Tikhonov). Specifically, we train a deep convolutional neural network (CNN) denoiser to generate a flexible deep denoiser prior to handle high noise. Then, we plug the deep denoiser prior as a modular part into a novel iterative optimization algorithm to handle blurring effects and propose an adaptive parameter selection strategy for the iterative optimization algorithm. Main results. Simulation results show that the deep denoiser prior plays the role of reducing noise intensity, while the novel iterative optimization algorithm and adaptive parameter selection strategy can effectively eliminate the semi-convergence property. They enable DP-RI-Tikhonov to achieve an average quantitative result (normalized root mean square error, structural similarity) of (0.1364, 0.9574) at the stopping iteration, outperforming a conventional PET IR method with an average quantitative result of (0.1533, 0.9523) and a state-of-the-art deep plug-and-play IR method with an average quantitative result of (0.1404, 0.9554). Moreover, the advantage of DP-RI-Tikhonov becomes more obvious at the last iteration. Experiments on six clinical whole-body PET images further indicate that DP-RI-Tikhonov successfully reduces noise intensity and recovers fine details, recovering sharper and more uniform images than the comparison methods. Significance. DP-RI-Tikhonov's ability to reduce noise intensity and effectively eliminate the semi-convergence property overcomes the limitations of existing methods. This advancement may have substantial implications for other medical IR.

show abstract

Section: Application To Simulation Data (1) Experimental Settingmentioning

confidence: 99%

Section: Application To Real Data (1) Experimental Settingmentioning

confidence: 99%