Zero-TE MRI-based attenuation correction for bone components on chest [<sup>18</sup>F] FDG PET/MRI: accuracy, repeatability, and external validation of an unsupervised deep learning approach using unpaired PET/CT data.
Authors
Affiliations (7)
Affiliations (7)
- Department of Radiology, Kobe University Graduate School of Medicine, 7- 5-2 Kusunoki-cho, Chuo-ku, Kobe, 650-0017, Hyogo, Japan. [email protected].
- Division of Medical Imaging, Biomedical Imaging Research Center, University of Fukui, Fukui, Japan. [email protected].
- Department of Radiology, Kobe University Graduate School of Medicine, 7- 5-2 Kusunoki-cho, Chuo-ku, Kobe, 650-0017, Hyogo, Japan.
- GE HealthCare, Munich, Germany.
- GE HealthCare, Hino, Japan.
- Department of Nuclear Medicine, University Hospital Zurich, University of Zurich, Zurich, Switzerland.
- Division of Medical Imaging, Biomedical Imaging Research Center, University of Fukui, Fukui, Japan.
Abstract
In positron emission tomography (PET)/magnetic resonance imaging (MRI), attenuation correction (AC) for PET of the head is achieved by MRI data to generate pseudo-computed tomography (CT) images. However, for the torso, AC becomes more challenging due to the complexity of separating bone components. Additionally, generating accurate MRI-based CT using deep learning poses significant difficulties for the chest, primarily because perfectly paired MRI and CT training data are hard to obtain owing to respiratory motion and body movements. We previously demonstrated that MRI-to-CT conversion can be achieved without deformation, even using unsupervised learning for zero echo time (ZTE) MRI and CT data from different individuals. Building on this foundation, our study aims to apply this approach to AC in chest PET/MRI and assess their quantitative accuracy, reproducibility, and external validity. The datasets used included (1) training dataset (unpaired ZTE MRI and CT of PET/CT, n = 360 and 500, respectively); (2) test dataset (paired PET/MRI and PET/CT, n = 25 and 25, respectively); (3) repeatability assessment dataset (repeated PET/MRI, n = 15 × 2 scans for the same patient); and (4) external validation dataset (paired MRI component of PET/MRI and CT, n = 30 and 30, respectively, acquired at another institution). Unpaired training data were used to train the deep learning model of pseudo-CT generation from ZTE. The accuracy, repeatability, and reproducibility of the PET/MRI scans using ZTE- and deep learning-based AC (MRAC<sub>ZTE</sub>) were evaluated based on the similarity of the histograms and the mean standardized uptake value (SUVmean) of physiological background of bone and liver. The histogram correlation coefficients between MRAC<sub>ZTE</sub> and the AC map based on the CT (CTAC) for the spine were significantly higher than those between conventional AC (MRAC<sub>Dixon</sub>) and CTAC. Additionally, bone SUVmean obtained using MRAC<sub>ZTE</sub> showed reduced bias relative to CTAC compared with MRAC<sub>Dixon</sub>. This method proved to be reproducible on each patient level and robust against external validation. Unsupervised learning with unpaired ZTE and CT data enabled pseudo-CT generation with bone components that closely matched CT-based attenuation maps. Integration into MR-based attenuation correction resulted in stable physiological uptake measurements in chest PET/MRI, supporting the feasibility of this approach.