<sup>18</sup>F-FDG PET-based liver segmentation using deep-learning.
Authors
Affiliations (7)
Affiliations (7)
- Department of Radiology, Fukushima Medical University Hospital, 1 Hikarigaoka, Fukushima, Fukushima, 960-1247, Japan.
- Department of Radiological Sciences, School of Health Sciences, Fukushima Medical University, 10-6 Sakaemachi, Fukushima-shi, Fukushima, 960- 8516, Japan.
- Department of Radiological Sciences, School of Health Sciences, Fukushima Medical University, 10-6 Sakaemachi, Fukushima-shi, Fukushima, 960- 8516, Japan. [email protected].
- Department of Molecular Imaging and Theranostics, National Institute of Radiological Sciences (NIRS), National Institutes for Quantum and Radiological Science and Technology (QST), 4-9-1, Anagawa, Inage-ku, Chiba, 263-0024, Japan. [email protected].
- Department of Molecular Imaging and Theranostics, National Institute of Radiological Sciences (NIRS), National Institutes for Quantum and Radiological Science and Technology (QST), 4-9-1, Anagawa, Inage-ku, Chiba, 263-0024, Japan.
- Biomedical Imaging Sciences, Department of Integrated Health Sciences, Nagoya University Graduate School of Medicine, Tokai National Education and Research System, Higashi-ku, Nagoya, Japan.
- Department of Radiological Sciences, International University of Health and Welfare, 2600-1 Kitakanemaru, Otawara, Tochigi, Japan.
Abstract
Organ segmentation using <sup>18</sup>F-FDG PET images alone has not been extensively explored. Segmentation based methods based on deep learning (DL) have traditionally relied on CT or MRI images, which are vulnerable to alignment issues and artifacts. This study aimed to develop a DL approach for segmenting the entire liver based solely on <sup>18</sup>F-FDG PET images. We analyzed data from 120 patients who were assessed using <sup>18</sup>F-FDG PET. A three-dimensional (3D) U-Net model from nnUNet and preprocessed PET images served as DL and input images, respectively, for the model. The model was trained with 5-fold cross-validation on data from 100 patients, and segmentation accuracy was evaluated on an independent test set of 20 patients. Accuracy was assessed using Intersection over Union (IoU), Dice coefficient, and liver volume. Image quality was evaluated using mean (SUVmean) and maximum (SUVmax) standardized uptake value and signal-to-noise ratio (SNR). The model achieved an average IoU of 0.89 and an average Dice coefficient of 0.94 based on test data from 20 patients, indicating high segmentation accuracy. No significant discrepancies in image quality metrics were identified compared with ground truth. Liver regions were accurately extracted from <sup>18</sup>F-FDG PET images which allowed rapid and stable evaluation of liver uptake in individual patients without the need for CT or MRI assessments.