Accuracy evaluation of liver and tumor auto-segmentation in CT images using 2D CoordConv DeepLab V3+ model in radiotherapy
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Medical image segmentation is the most important task in radiation therapy. Especially, when segmenting medical images, the liver is one of the most difficult organs to segment because it has various shapes and is close to other organs. Therefore, automatic segmentation of the liver in computed tomography (CT) images is a difficult task. Since tumors also have low contrast in surrounding tissues, and the shape, location, size, and number of tumors vary from patient to patient, accurate tumor segmentation takes a long time. In this study, we propose a method algo- rithm for automatically segmenting the liver and tumor for this purpose. As an advantage of setting the boundaries of the tumor, the liver and tumor were automatically segmented from the CT image using the 2D CoordConv Dee- pLab V3+ model using the CoordConv layer. For tumors, only cropped liver images were used to improve accuracy. Additionally, to increase the segmentation accuracy, augmentation, preprocess, loss function, and hyperparameter were used to find optimal values. We compared the CoordConv DeepLab v3+ model using the CoordConv layer and the DeepLab V3+ model without the CoordConv layer to determine whether they affected the segmentation accu- racy. The data sets used included 131 hepatic tumor segmentation (LiTS) challenge data sets (100 train sets, 16 val- idation sets, and 15 test sets). Additional learned data were tested using 15 clinical data from Seoul St. Mary's Hospital. The evaluation was compared with the study results learned with a two-dimensional deep learning-based model. Dice values without the CoordConv layer achieved 0.965 ± 0.01 for liver segmentation and 0.925 ± 0.04 for tumor segmentation using the LiTS data set. Results from the clinical data set achieved 0.927 ± 0.02 for liver division and 0.903 ± 0.05 for tumor division. The dice values using the CoordConv layer achieved 0.989 ± 0.02 for liver seg- mentation and 0.937 ± 0.07 for tumor segmentation using the LiTS data set. Results from the clinical data set achieved 0.944 ± 0.02 for liver division and 0.916 ± 0.18 for tumor division. The use of CoordConv layers improves the segmentation accuracy. The highest of the most recently published values were 0.960 and 0.749 for liver and tumor division, respectively. However, better performance was achieved with 0.989 and 0.937 results for liver and tumor, which would have been used with the algorithm proposed in this study. The algorithm proposed in this study can play a useful role in treatment planning by improving contouring accuracy and reducing time when segmentation evaluation of liver and tumor is performed. And accurate identification of liver anatomy in medical imaging appli- cations, such as surgical planning, as well as radiotherapy, which can leverage the findings of this study, can help clin- ical evaluation of the risks and benefits of liver intervention.
Abstract
Medical image segmentation is the most important task in radiation therapy. Especially, when segmenting medical images, the liver is one of the most difficult organs to segment because it has various shapes and is close to other organs. Therefore, automatic segmentation of the liver in computed tomography (CT) images is a difficult task. Since tumors also have low contrast in surrounding tissues, and the shape, location, size, and number of tumors vary from patient to patient, accurate tumor segmentation takes a long time. In this study, we propose a method algo- rithm for automatically segmenting the liver and tumor for this purpose. As an advantage of setting the boundaries of the tumor, the liver and tumor were automatically segmented from the CT image using the 2D CoordConv Dee- pLab V3+ model using the CoordConv layer. For tumors, only cropped liver images were used to improve accuracy. Additionally, to increase the segmentation accuracy, augmentation, preprocess, loss function, and hyperparameter were used to find optimal values. We compared the CoordConv DeepLab v3+ model using the CoordConv layer and the DeepLab V3+ model without the CoordConv layer to determine whether they affected the segmentation accu- racy. The data sets used included 131 hepatic tumor segmentation (LiTS) challenge data sets (100 train sets, 16 val- idation sets, and 15 test sets). Additional learned data were tested using 15 clinical data from Seoul St. Mary's Hospital. The evaluation was compared with the study results learned with a two-dimensional deep learning-based model. Dice values without the CoordConv layer achieved 0.965 ± 0.01 for liver segmentation and 0.925 ± 0.04 for tumor segmentation using the LiTS data set. Results from the clinical data set achieved 0.927 ± 0.02 for liver division and 0.903 ± 0.05 for tumor division. The dice values using the CoordConv layer achieved 0.989 ± 0.02 for liver seg- mentation and 0.937 ± 0.07 for tumor segmentation using the LiTS data set. Results from the clinical data set achieved 0.944 ± 0.02 for liver division and 0.916 ± 0.18 for tumor division. The use of CoordConv layers improves the segmentation accuracy. The highest of the most recently published values were 0.960 and 0.749 for liver and tumor division, respectively. However, better performance was achieved with 0.989 and 0.937 results for liver and tumor, which would have been used with the algorithm proposed in this study. The algorithm proposed in this study can play a useful role in treatment planning by improving contouring accuracy and reducing time when segmentation evaluation of liver and tumor is performed. And accurate identification of liver anatomy in medical imaging appli- cations, such as surgical planning, as well as radiotherapy, which can leverage the findings of this study, can help clin- ical evaluation of the risks and benefits of liver intervention.
참고문헌 (25)
- Gérard A, "Preopera-tive radiotherapy as adjuvant treatment in rectal cancer. Final results of a randomized study of the European Organization for Research and Treatment of Cancer (EORTC)", Annals of surgery, 208(5), pp.606~614 (1988)
- Lei Chen, "Liver tumor segmentation in CT vol-umes using an adversarial densely connected network", BMC bioinformatics, 20, pp.1~13 (2019)
- Yuka Urago, "Evaluation of auto-segmentation accuracy of cloud-based artificial intelligence and atlas-based models", Radiation Oncology, 16(1), pp.1~13 (2021)
- Llovet JM, "Hepatocellular carcinoma", Nature reviews. Disease primers, 7(1), pp.6~ (2021)
- Hyunseok Seo, "Modified U-Net (mU-Net) with incorporation of object-dependent high level features for improved liver and liver-tumor segmentation in CT images", IEEE transactions on medical imaging, 39(5), pp.1316~1325 (2019)
- Jean-François Daisne, "Atlas-based automatic segmentation of head and neck organs at risk and nodal target volumes: a clinical validation", Radiation oncology, 8(1), pp.1~11 (2013)
- Shalini K. Vinod, "Holloway. Uncertainties in volume delineation in radiation oncol-ogy: a systematic review and recommendations for future studies", Radiotherapy and Oncology, 121(2), pp.169~179 (2016)
- Elif DeryaÜbeyli, "Implementing automated diagnostic sys-tems for breast cancer detection", Expert systems with Appli-cations, 33(4), pp.1054~1062 (2007)
- Berkman Sahiner, "Deep learning in medical imaging and radiation therapy", Medical physics, 46(1), pp.1~36 (2019)
- Jonathan Long, "Fully con-volutional networks for semantic segmentation", Proceedings of the IEEE conference on computer vision and pattern rec-ognition, pp.3431~3440 (2015)
- Holger R. Roth, "An application of cascaded 3D fully convolutional networks for medical image segmentation", Computerized Medical Imaging and Graphics, 66, pp.90~99 (2018)
- Patrick Ferdinand Christ, "Automatic liver and lesion segmentation in CT using cascaded fully convolutional neural networks and 3D conditional ran-dom fields", International conference on medical image comput-ing and computer-assisted intervention, 9901, pp.415~423 (2016)
- Peijun Hu, "Automatic abdominal multi-organ segmentation using deep convolutional neural network and time-implicit level sets", International journal of computer assisted radiology and surgery, 12(3), pp.399~411 (2017)
- Meg F. Bobo, "Fully convolutional neural networks improve abdominal organ segmentation", Medical Imaging, 10574, pp.750~757 (2018)
- Tongle Fan, "Ma-net:A multi-scale attention network for liver and tumor segmen-tation", IEEE Access, 8, pp.179656~179665 (2020)
- Nasser Alalwan, "Efficient 3D deep learning model for medical image semantic segmentation", Alexandria Engineering Journal, 60(1), pp.1231~1239 (2021)
- Amit Pandey, "Comparative analysis of KNN algorithm using various normalization techniques", International Journal of Computer Network and Information Security, 9(11), pp.36~ (2017)
- Marcin Kociołek, "Does image normalization and intensity resolution impact texture classification?", Computerized Medical Imaging and Graph-ics, 81, pp.101716~ (2020)
- El Jurdi R, "CoordConv-Unet: Investigating CoordConv for Organ Segmentation", IRBM, 42(6), pp.415~423 (2021)
- Tsung-Yi Lin, "Focal loss for dense object detection", Proceedings of the IEEE international conference on computer vision, pp.2980~2988 (2017)
- Seyed Sadegh Mohseni Salehi, "Tversky loss function for image segmentation using 3D fully convolutional deep networks", International workshop on machine learning in medical imaging, 10541, pp.379~387 (2017)