Evaluation of data-driven respiratory gating for subcentimeter lesions using digital PET/CT system and three-axis
Shohei Fukai1,2, Hiromitsu Daisaki2,3, Naoki Shimada1
1Department of Nuclear Medicine, The Cancer Institute Hospital of Japanese Foundation for Cancer Research, 3-8-31 Ariake, Koto-ku, Tokyo 135-8550, Japan.
Abstract:
Introduction.The application of data-driven respiratory gating (DDG) for subcentimeter lesions with respiratory movement remains poorly understood. Hence, this study aimed to clarify DDG application for subcentimeter lesions and the ability of digital Positron emission tomography/computed tomography (PET/CT) system combined with DDG to detect these lesions under three-axis respiration.Methods.Discovery MI PET/CT system and National Electrical Manufacturers Association (NEMA) body phantom with Micro Hollow Sphere (4, 5, 6, 8, 10, and 13 mm) were used. The NEMA phantom was filled with18F-FDG solutions of 42.4 and 5.3 kBq/ml for each hot sphere and background region. The 3.6 s cycles of three-axis respiratory motion were reproduced using the motion platform UniTraQ. The PET data acquisition was performed in stationary and respiratory-moving states. The data were reconstructed in three PET groups: stationary (NM-PET), no gating with respiratory movement (NG-PET), and DDG gating with respiratory movement (DDG-PET) groups. For image quality, percent contrast (QH); maximum, peak, and mean standardized uptake value (SUV); background region; and detectability index (DI) were evaluated in each PET group. Visual assessment was also conducted.Results.The groups with respiratory movement had deteriorated QHand SUVs compared with NM-PET. Compared with NG-PET, DDG-PET has significantly improved QHand SUVs in spheres above 6 mm. The background region showed no significant difference between groups. The SUVmax, SUVpeak, and QHvalues of 8 mm sphere were highest in NM-PET, followed by DDG-PET and NG-PET. In visual assessment, the spheres above 6 mm were detected in all PET groups. DDG application did not detect new lesions, but it increased DI and visual score.Conclusions. The application of principal component analysis (PCA)-based DDG algorithm improves both image quality and quantitative SUVs in subcentimeter lesions measuring above 6 mm. Although DDG application cannot detect new subcentimeter lesions, it increases the visual indices.
