Automatic pancreatic cancer GTV segmentation based on deep learning(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2025年第7期
- Page:
- 923-928
- Research Field:
- 医学影像物理
- Publishing date:
Info
- Title:
- Automatic pancreatic cancer GTV segmentation based on deep learning
- Author(s):
- CHEN Chaoshuang1; CAO Yangsen2; ZHU Xiaofei2; ZENG Fubin2; GU Lei2; JIANG Lingong2; ZHANG Huojun2
- 1. Department of Nuclear Medicine and Oncology, Chinese People’s Liberation Army 92493 Military Hospital, Huludao 125000, China;2. Department of Radiation Oncology, Changhai Hospital of Naval Medical University, Shanghai 200433, China
- Keywords:
- pancreatic cancer; gross tumor volume; segmentation; deep learning; convolutional neural network
- PACS:
- R318.14
- DOI:
- DOI:10.3969/j.issn.1005-202X.2025.07.012
- Abstract:
- Abstract: Objective To investigate the feasibility and accuracy of convolutional neural networks for automaticallydelineating the pancreatic cancer gross target volume (GTV) in pancreatic enhanced CT. Methods The localizable enhancedCT images of 114 patients with pancreatic cancer were retrospectively selected, in which the GTV was manually delineatedusing AccuContour. The imaging data were then import to AccuLearning and randomly divided as the training set, validationset and test set at a ratio of 8:1:1. Flex and Segresnet were used to train the automatic segmentation model, with each networkstructure trained continuously 3 times using fixed training parameters. The model was evaluated in terms of Dice similaritycoefficient (DSC), 95% Hausdorff distance (HD95), average symmetric surface distance (ASSD) and relative volumedifference (RVD). Results In the model training phase, Flex-3 test results in Flex group were the worst, with a minimum DSCof 0.14% and an average DSC of 56.30%, while Flex-1 performed well, achieving a minimum DSC of 47.90% and anaverage DSC of 67.35%. Meanwhile, Segresnet-2 in Segresnet group had the worst test results, with a minimum DSC of0.00% and an average DSC of 42.46%, while Segresnet-3 test results were better, with a minimum DSC of 42.65% and anaverage DSC of 63.28%. In the fixed testing phase, the best results among all were as follows: average DSC and RVD valuesof 63.88% and 29.41% in Segresnet-3 group, average ASSD value of 4.43 mm in Segresnet-2 group, and average HD95 valueof 12.87 mm in Segresnet-1 group. Conclusion Both Flex and Segresnet architectures of convolutional neural network can beused for the automatic pancreatic tumor GTV segmentation training, with Segresnet performing better in comprehensiveevaluation.
Last Update: 2025-07-25