Development of Novel Residual-Dense-Attention (RDA) U-Net Network Architecture for Hepatocellular Carcinoma Segmentation

被引:9
作者
Chen, Wen-Fan [1 ]
Ou, Hsin-You [2 ,3 ,4 ]
Lin, Han-Yu [5 ]
Wei, Chia-Po [6 ]
Liao, Chien-Chang [2 ,3 ,4 ]
Cheng, Yu-Fan [2 ,3 ,4 ]
Pan, Cheng-Tang [5 ,7 ]
机构
[1] Natl Sun Yat Sen Univ, Inst Med Sci & Technol, Kaohsiung 80424, Taiwan
[2] Kaohsiung Chang Gung Mem Hosp, Liver Transplantat Program, Kaohsiung 83301, Taiwan
[3] Kaohsiung Chang Gung Mem Hosp, Dept Diagnost Radiol, Kaohsiung 83301, Taiwan
[4] Kaohsiung Chang Gung Mem Hosp, Surg, Kaohsiung 83301, Taiwan
[5] Natl Sun Yat Sen Univ, Dept Mech & Electromech Engn, Kaohsiung 80424, Taiwan
[6] Natl Sun Yat Sen Univ, Dept Elect Engn, Kaohsiung 80424, Taiwan
[7] Natl Sun Yat Sen Univ, Coll Semicond & Adv Technol Res, Inst Adv Semicond Packaging & Testing, Kaohsiung 80424, Taiwan
关键词
computed tomography; hepatocellular carcinoma; attention U-Net; ResNet; DenseNet; staging classification;
D O I
10.3390/diagnostics12081916
中图分类号
R5 [内科学];
学科分类号
1002 ; 100201 ;
摘要
The research was based on the image recognition technology of artificial intelligence, which is expected to assist physicians in making correct decisions through deep learning. The liver dataset used in this study was derived from the open source website (LiTS) and the data provided by the Kaohsiung Chang Gung Memorial Hospital. CT images were used for organ recognition and lesion segmentation; the proposed Residual-Dense-Attention (RDA) U-Net can achieve high accuracy without the use of contrast. In this study, U-Net neural network was used to combine ResBlock in ResNet with Dense Block in DenseNet in the coder part, allowing the training to maintain the parameters while reducing the overall recognition computation time. The decoder was equipped with Attention Gates to suppress the irrelevant areas of the image while focusing on the significant features. The RDA model was used to identify and segment liver organs and lesions from CT images of the abdominal cavity, and excellent segmentation was achieved for the liver located on the left side, right side, near the heart, and near the lower abdomen with other organs. Better recognition was also achieved for large, small, and single and multiple lesions. The study was able to reduce the overall computation time by about 28% compared to other convolutions, and the accuracy of liver and lesion segmentation reached 96% and 94.8%, with IoU values of 89.5% and 87%, and AVGDIST of 0.28 and 0.80, respectively.
引用
收藏
页数:15
相关论文
共 37 条
[1]   On the usage of average Hausdorff distance for segmentation performance assessment: hidden error when used for ranking [J].
Aydin, Orhun Utku ;
Taha, Abdel Aziz ;
Hilbert, Adam ;
Khalil, Ahmed A. ;
Galinovic, Ivana ;
Fiebach, Jochen B. ;
Frey, Dietmar ;
Madai, Vince Istvan .
EUROPEAN RADIOLOGY EXPERIMENTAL, 2021, 5 (01)
[2]   SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation [J].
Badrinarayanan, Vijay ;
Kendall, Alex ;
Cipolla, Roberto .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2017, 39 (12) :2481-2495
[3]  
Carbonell J.G., 2018, PROC 6 INT C REPRESE
[4]  
codalab, LITS CHALLENGE
[5]   MRI Breast Tumor Segmentation Using Different Encoder and Decoder CNN Architectures [J].
El Adoui, Mohammed ;
Mahmoudi, Sidi Ahmed ;
Larhmam, Mohamed Amine ;
Benjelloun, Mohammed .
COMPUTERS, 2019, 8 (03)
[6]   Machine Learning for Medical Imaging1 [J].
Erickson, Bradley J. ;
Korfiatis, Panagiotis ;
Akkus, Zeynettin ;
Kline, Timothy L. .
RADIOGRAPHICS, 2017, 37 (02) :505-515
[7]   A guide to deep learning in healthcare [J].
Esteva, Andre ;
Robicquet, Alexandre ;
Ramsundar, Bharath ;
Kuleshov, Volodymyr ;
DePristo, Mark ;
Chou, Katherine ;
Cui, Claire ;
Corrado, Greg ;
Thrun, Sebastian ;
Dean, Jeff .
NATURE MEDICINE, 2019, 25 (01) :24-29
[8]  
Ethem Alpaydin, 2020, INTRO MACHINE LEARNI
[9]  
Gelband H., 2015, LIVER CANC CANCER DI, V3
[10]   Machine Learning in Medical Imaging [J].
Giger, Maryellen L. .
JOURNAL OF THE AMERICAN COLLEGE OF RADIOLOGY, 2018, 15 (03) :512-520