Hybrid Multiple-Organ Segmentation Method Using Multiple U-Nets in PET/CT Images
Source abstract
Abstract
PET/CT can scan low-dose computed tomography (LDCT) images with morphological information and PET images with functional information. Because the whole body is targeted for imaging, PET/CT examinations are important in cancer diagnosis. However, the several images obtained by PET/CT place a heavy burden on radiologists during diagnosis. Thus, the development of computer-aided diagnosis (CAD) and technologies assisting in diagnosis has been requested. However, because FDG accumulation in PET images differs for each organ, recognizing organ regions is essential for developing lesion detection and analysis algorithms for PET/CT images. Therefore, we developed a method for automatically extracting organ regions from PET/CT images using U-Net or DenseUNet, which are deep-learning-based segmentation networks. The proposed method is a hybrid approach combining morphological and functional information obtained from LDCT and PET images. Moreover, pre-training using ImageNet and RadImageNet was performed and compared. The best extraction accuracy was obtained by pre-training ImageNet with Dice indices of 94.1, 93.9, 91.3, and 75.1% for the liver, kidney, spleen, and pancreas, respectively. This method obtained better extraction accuracy for low-quality PET/CT images than did existing studies on PET/CT images and was comparable to existing studies on diagnostic contrast-enhanced CT images using the hybrid method and pre-training.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Radiomics and Machine Learning in Medical Imaging · Medical Imaging Techniques and Applications · Advanced X-ray and CT Imaging
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Graduate School of Health Sciences, Fujita Health University, Toyoake 470-1192, Japan
Faculty of Information Engineering, Meijo University, Nagoya 468-8502, Japan
Faculty of Engineering, Gifu University, Gifu 501-1193, Japan
Department of Artificial Intelligence Diagnostic Radiology, Graduate School of Medicine, Osaka University, Osaka 565-0871, Japan
Department of Diagnostic and Interventional Radiology, Graduate School of Medicine, Osaka University, Osaka 565-0871, Japan
References (20)
Show source-provided references
1. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries10.3322/caac.21660 ↗
2. Automated detection of pulmonary nodules in PET/CT images: Ensemble false-positive reduction using a convolutional neural network technique10.1118/1.4948498 ↗
3. Lung cancer detection and classification with 3D convolutional neural network (3D-CNN)
4. Deep learning for fully-automated localization and segmentation of rectal cancer on multiparametric MR10.1038/s41598-017-05728-9 ↗
5. Deep learning in mammography images segmentation and classification: Automated CNN approach10.1016/j.aej.2021.03.048 ↗
6. Automated abdominal multi-organ segmentation with subject-specific atlas generation10.1109/TMI.2013.2265805 ↗
7. Discriminative dictionary learning for abdominal multi-organ segmentation10.1016/j.media.2015.04.015 ↗
8. Multi-organ localization with cascaded global-to-local regression and shape prior10.1016/j.media.2015.04.007 ↗
9. Regression forests for efficient anatomy detection and localization in computed tomography scans10.1016/j.media.2013.01.001 ↗
10. Automatic abdominal multi-organ segmentation using deep convolutional neural network and time-implicit level sets10.1007/s11548-016-1501-5 ↗
11. Deep learning of the sectional appearances of 3D CT images for anatomical structure segmentation based on an FCN voting method10.1002/mp.12480 ↗
12. Roth, H.R., Oda, H., Hayashi, Y., Oda, M., Shimizu, N., Fujiwara, M., Misawa, K., and Mori, K. (2017). Hierarchical 3D Fully Convolutional Networks for Multi-organ Segmentation. arXiv.10.1007/978-3-030-00937-3_48 ↗
13. Dual-modality multi-atlas segmentation of torso organs from [18F]FDG-PET/CT images10.1007/s11548-018-1879-3 ↗
14. Multiple organ-specific cancers classification from PET/CT images using deep learning10.1007/s11042-022-12055-3 ↗
15. ImageNet classification with deep convolutional neural networks10.1145/3065386 ↗
16. RadImageNet: An open radiologic deep learning research dataset for effective transfer learning10.1148/ryai.210315 ↗
17. U-net: Convolutional networks for biomedical image segmentation10.1007/978-3-319-24574-4_28 ↗
18. H-DenseUNet: Hybrid densely connected UNet for liver and tumor segmentation from CT volumes10.1109/TMI.2018.2845918 ↗
19. Combo loss: Handling Input and Output Imbalance in multi-organ Segmentation10.1016/j.compmedimag.2019.04.005 ↗
20. Automatic multi-organ segmentation on abdominal CT with dense V-networks10.1109/TMI.2018.2806309 ↗