基於電腦斷層掃描的自動肝臟腫瘤檢測和分割可以幫助臨床檢查更快、更準確地診斷潛在病變。然而，深度學習檢測算法具有高靈敏度和低精度的特點，這使得自動檢測系統在能夠準確地減少偵測的偽陽性腫瘤數量之前，無法有效加快診斷過程。檢測模型產生偽陽性腫瘤的主要原因是由於缺乏從全局角度學習肝臟結構知識的能力，這使得模型將部分體積偽影視為病變。針對上述限制，本研究提出了一種新穎的切片融合方法，透過挖掘目標斷層掃描切片中組織之間的全局結構關係來獲得空間注意力圖，並根據組織的重要性來融合相鄰斷層掃描切片的特徵信息。我們使用 Mask R-CNN 作為基本檢測架構，並在公開的 MICCAI 2017 Liver in Tumor Segmentation (LiTS) Challenge 數據集和國立成功大學醫院數據集上評估所提出的方法。大量實驗表明，所提出的方法不僅可以通過減少所偵測小於十毫米的假病灶數量來提高模型的腫瘤檢測能力，同時方法也提升了模型的腫瘤分割性能。沒有使用花里胡哨的技巧，本研究提出的單一模型在 LiTS 測試數據集上的肝腫瘤檢測和分割與其他最先進的模型相比表現出優越的性能。

Automatic liver tumor detection and segmentation based on computed tomography (CT) can help clinical examinations diagnose potential lesions faster and more accurately. However, deep learning-based detection algorithms have the characteristics of high Sensitivity and low Precision. This fact makes the automatic detection system unable to effectively speed up the diagnosis process until it can accurately reduce the number of false-positive detected tumors. The main reason that the detection model produces false-positive tumors is due to the lack of the ability to learn the knowledge of the liver structure from a global perspective, which makes model regards partial volume artifacts as lesions. To address above limitation, this research proposes a novel slice-fusion method, which obtains the spatial attention map by mining the global structural relationship between the tissues in the target CT slices and fuses the features of adjacent slices according to the importance of the tissues. We use Mask R-CNN as the basic detection architecture and evaluate the proposed method on the public MICCAI 2017 Liver in Tumor Segmentation (LiTS) Challenge dataset and the external National Cheng Kung University Hospital (NCKUH) dataset. Extensive experiments have demonstrated that proposed method can not only enhance tumor detection ability via reducing the number of pseudo-lesions smaller than 10 mm, but also improve performance of tumor segmentation. Without bells and whistles, the single proposed detection model has shown outstanding performance in liver tumor detection and segmentation on LiTS test dataset compared with other state-of-the-art models.

[ 1 ] C. Van de Velde, “Treatment of liver metastases of colorectal cancer,” Annals of oncology, vol. 16, pp. ii144–ii149, 2005.
[ 2 ] R. Adam, V. Delvart, G. Pascal, A. Valeanu, D. Castaing, D. Azoulay, S. Giacchetti, B. Paule, F. Kunstlinger, O. Gh´emard, et al., “Rescue surgery for unresectable colorectal liver metastases downstaged by chemotherapy: a model to predict long-term survival,” Annals of surgery, vol. 240, no. 4, p. 644, 2004.
[ 3 ] S. Manfredi, C. Lepage, C. Hatem, O. Coatmeur, J. Faivre, and A.-M. Bouvier, “Epidemiology and management of liver metastases from colorectal cancer,” Annals of surgery, vol. 244, no. 2, p. 254, 2006.
[ 4 ] A. C. Wei, P. D. Greig, D. Grant, B. Taylor, B. Langer, and S. Gallinger,“Survival after hepatic resection for colorectal metastases: a 10-year experience,” Annals of surgical oncology, vol. 13, no. 5, pp. 668–676, 2006.
[ 5 ] M. G. House, H. Ito, M. G¨onen, Y. Fong, P. J. Allen, R. P. DeMatteo, M. F. Brennan, L. H. Blumgart, W. R. Jarnagin, and M. I. D’Angelica,“Survival after hepatic resection for metastatic colorectal cancer: trends in outcomes for 1,600 patients during two decades at a single institution,” Journal of the American College of Surgeons, vol. 210, no. 5, pp. 744–752, 2010.
[ 6 ] J. N. Leal, A. K. Bressan, N. Vachharajani, M. Gonen, T. P. Kingham, M. I. D’Angelica, P. J. Allen, R. P. DeMatteo, M. B. Doyle, O. F. Bathe, et al., “Time-to-surgery and survival outcomes in resectable colorectal liver metastases: a multi-institutional evaluation,” Journal of the American College of Surgeons, vol. 222, no. 5, pp. 766–779, 2016.
[ 7 ] K. He, G. Gkioxari, P. Doll´ar, and R. Girshick, “Mask r-cnn,” in Proceedings of the IEEE international conference on computer vision, pp. 2961–2969, 2017.
[ 8 ] J. Redmon and A. Farhadi, “Yolov3: An incremental improvement,” arXiv preprint arXiv:1804.02767, 2018.
[ 9 ] T.-Y. Lin, P. Goyal, R. Girshick, K. He, and P. Doll´ar, “Focal loss for dense object detection,” in Proceedings of the IEEE international conference on computer vision, pp. 2980–2988, 2017.
[ 10 ] R. Zhang, C. Cheng, X. Zhao, and X. Li, “Multiscale mask r-cnn–based lung tumor detection using pet imaging,” Molecular imaging, vol. 18, p. 1536012119863531, 2019.
[ 11 ] H. Gamage, W. Wijesinghe, and I. Perera, “Instance-based segmentation for boundary detection of neuropathic ulcers through mask-rcnn,” in International Conference on Artificial Neural Networks, pp. 511–522, Springer, 2019.
[ 12 ] R. Anantharaman, M. Velazquez, and Y. Lee, “Utilizing mask r-cnn for detection and segmentation of oral diseases,” in 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 2197–2204, IEEE, 2018.
[ 13 ] J. George, S. Skaria, V. Varun, et al., “Using yolo based deep learning network for real time detection and localization of lung nodules from low dose ct scans,” in Medical Imaging 2018: Computer-Aided Diagnosis, vol. 10575, p. 105751I, International Society for Optics and Photonics, 2018.
[ 14 ] S. Pang, T. Ding, S. Qiao, F. Meng, S. Wang, P. Li, and X. Wang, “A novel yolov3-arch model for identifying cholelithiasis and classifying gallstones on ct images,” PloS one, vol. 14, no. 6, p. e0217647, 2019.
[ 15 ] M. Zlocha, Q. Dou, and B. Glocker, “Improving retinanet for ct lesion detection with dense masks from weak recist labels,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 402–410, Springer, 2019.
[ 16 ] I. W. Harsono, S. Liawatimena, and T. W. Cenggoro, “Lung nodule detection and classification from thorax ct-scan using retinanet with transfer learning,” Journal of King Saud University-Computer and Information Sciences, 2020.
[ 17 ] K. Kim, S. Kim, K. Han, H. Bae, J. Shin, and J. S. Lim, “Diagnostic performance of deep learning-based lesion detection algorithm in ct for detecting hepatic metastasis from colorectal cancer,” Korean journal of radiology, vol. 22, no. 6, p. 912, 2021.
[ 18 ] O. J. Garden, M. Rees, G. Poston, D. Mirza, M. Saunders, J. Ledermann, J. Primrose, and R. Parks, “Guidelines for resection of colorectal cancer liver metastases,” Gut, vol. 55, no. suppl 3, pp. iii1–iii8, 2006.
[ 19 ] D. W. Kim, G. Lee, S. Y. Kim, G. Ahn, J.-G. Lee, S. S. Lee, K. W. Kim, S. H. Park, Y. J. Lee, and N. Kim, “Deep learning–based algorithm to detect primary hepatic malignancy in multiphase ct of patients at high risk for hcc,” European Radiology, pp. 1–11, 2021.
[ 20 ] R. Chou, C. Cuevas, R. Fu, B. Devine, N. Wasson, A. Ginsburg, B. Zakher, M. Pappas, E. Graham, and S. D. Sullivan, “Imaging techniques for the diagnosis of hepatocellular carcinoma: a systematic review and meta-analysis,” Annals of internal medicine, vol. 162, no. 10, pp. 697–711, 2015.
[ 21 ] L. R. Roberts, C. B. Sirlin, F. Zaiem, J. Almasri, L. J. Prokop, J. K. Heimbach, M. H. Murad, and K. Mohammed, “Imaging for the diagnosis of hepatocellular carcinoma: A systematic review and meta-analysis,” Hepatology, vol. 67, no. 1, pp. 401–421, 2018.
[ 22 ] J. F. Barrett and N. Keat, “Artifacts in ct: recognition and avoidance,” Radiographics, vol. 24, no. 6, pp. 1679–1691, 2004.
[ 23 ] B. L. Triche, J. T. Nelson Jr, N. S. McGill, K. K. Porter, R. Sanyal, F. N. Tessler, J. E. McConathy, D. M. Gauntt, M. V. Yester, and S. P. Singh,“Recognizing and minimizing artifacts at ct, mri, us, and molecular imaging,” RadioGraphics, vol. 39, no. 4, pp. 1017–1018, 2019.
[ 24 ] Z. Zhang, C. Lan, W. Zeng, X. Jin, and Z. Chen, “Relation-aware global attention for person re-identification,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 3186–3195, 2020.
[ 25 ] A. Sinha and J. Dolz, “Multi-scale self-guided attention for medical image segmentation,” IEEE journal of biomedical and health informatics, 2020.
[ 26 ] N. Abraham and N. M. Khan, “A novel focal tversky loss function with improved attention u-net for lesion segmentation,” in 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019), pp. 683–687, IEEE, 2019.
[ 27 ] A. Vakanski, M. Xian, and P. E. Freer, “Attention-enriched deep learning model for breast tumor segmentation in ultrasound images,” Ultrasound in Medicine & Biology, vol. 46, no. 10, pp. 2819–2833, 2020.
[ 28 ] H. Zhang, J. Zhang, Q. Zhang, J. Kim, S. Zhang, S. A. Gauthier, P. Spincemaille, T. D. Nguyen, M. Sabuncu, and Y. Wang, “Rsanet: Recurrent slice-wise attention network for multiple sclerosis lesion segmentation,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 411–419, Springer, 2019.
[ 29 ] X. Wang, S. Han, Y. Chen, D. Gao, and N. Vasconcelos, “Volumetric attention for 3d medical image segmentation and detection,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 175–184, Springer, 2019.
[ 30 ] R. Gu, G. Wang, T. Song, R. Huang, M. Aertsen, J. Deprest, S. Ourselin, T. Vercauteren, and S. Zhang, “Ca-net: Comprehensive attention convolutional neural networks for explainable medical image segmentation,” IEEE transactions on medical imaging, 2020.
[ 31 ] Q. Jin, Z. Meng, C. Sun, H. Cui, and R. Su, “Ra-unet: A hybrid deep attention-aware network to extract liver and tumor in ct scans,” Frontiers in Bioengineering and Biotechnology, vol. 8, p. 1471, 2020.
[ 32 ] F. Wang, M. Jiang, C. Qian, S. Yang, C. Li, H. Zhang, X. Wang, and X. Tang, “Residual attention network for image classification,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 3156–3164, 2017.
[ 33 ] H. Jiang, T. Shi, Z. Bai, and L. Huang, “Ahcnet: An application of attention mechanism and hybrid connection for liver tumor segmentation in ct volumes,” IEEE Access, vol. 7, pp. 24898–24909, 2019.
[ 34 ] O. Oktay, J. Schlemper, L. L. Folgoc, M. Lee, M. Heinrich, K. Misawa, K. Mori, S. McDonagh, N. Y. Hammerla, B. Kainz, et al., “Attention u-net: Learning where to look for the pancreas,” arXiv preprint arXiv:1804.03999, 2018.
[ 35 ] R. Hasegawa, Y. Iwamoto, H. Xianhua, L. Lin, H. Hu, C. Xiujun, and Y.-W. Chen, “Automatic detection and segmentation of liver tumors in multi-phase ct images by phase attention mask r-cnn,” in 2021 IEEE International Conference on Consumer Electronics (ICCE), pp. 1–5, IEEE, 2021.
[ 36 ] W. Luo, Y. Li, R. Urtasun, and R. Zemel, “Understanding the effective receptive field in deep convolutional neural networks,” in Proceedings of the 30th International Conference on Neural Information Processing Systems, pp. 4905–4913, 2016.
[ 37 ] X. Wang, R. Girshick, A. Gupta, and K. He, “Non-local neural networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 7794–7803, 2018.
[ 38 ] Y. Cao, J. Xu, S. Lin, F. Wei, and H. Hu, “Gcnet: Non-local networks meet squeeze-excitation networks and beyond,” in Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops, pp. 0–0, 2019.
[ 39 ] H.-C. Shin, H. R. Roth, M. Gao, L. Lu, Z. Xu, I. Nogues, J. Yao, D. Mollura, and R. M. Summers, “Deep convolutional neural networks for computer-aided detection: Cnn architectures, dataset characteristics and transfer learning,” IEEE transactions on medical imaging, vol. 35, no. 5, pp. 1285–1298, 2016.
[ 40 ] J. Yun, J. Park, D. Yu, J. Yi, M. Lee, H. J. Park, J.-G. Lee, J. B. Seo, and N. Kim, “Improvement of fully automated airway segmentation on volumetric computed tomographic images using a 2.5 dimensional convolutional neural net,” Medical image analysis, vol. 51, pp. 13–20, 2019.
[ 41 ] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778, 2016.
[ 42 ] S. Xie, R. Girshick, P. Doll´ar, Z. Tu, and K. He, “Aggregated residual transformations for deep neural networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1492–1500, 2017.
[ 43 ] T.-Y. Lin, P. Doll´ar, R. Girshick, K. He, B. Hariharan, and S. Belongie,“Feature pyramid networks for object detection,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 2117–2125, 2017.
[ 44 ] O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in International Conference on Medical image computing and computer-assisted intervention, pp. 234–241, Springer, 2015.
[ 45 ] J. Dai, H. Qi, Y. Xiong, Y. Li, G. Zhang, H. Hu, and Y.Wei, “Deformable convolutional networks,” in Proceedings of the IEEE international conference on computer vision, pp. 764–773, 2017.
[ 46 ] Z. Cai and N. Vasconcelos, “Cascade r-cnn: high quality object detection and instance segmentation,” IEEE transactions on pattern analysis and machine intelligence, 2019.
[ 47 ] C. R. Harris, K. J. Millman, S. J. van der Walt, R. Gommers, P. Virtanen, D. Cournapeau, E. Wieser, J. Taylor, S. Berg, N. J. Smith, et al., “Array programming with numpy,” Nature, vol. 585, no. 7825, pp. 357–362, 2020.
[ 48 ] B. Horn, B. Klaus, and P. Horn, Robot vision. MIT press, 1986.
[ 49 ] A. Paszke, S. Gross, S. Chintala, G. Chanan, E. Yang, Z. DeVito, Z. Lin, A. Desmaison, L. Antiga, and A. Lerer, “Automatic differentiation in pytorch,” 2017.
[ 50 ] K. Chen, J. Wang, J. Pang, Y. Cao, Y. Xiong, X. Li, S. Sun, W. Feng, Z. Liu, J. Xu, et al., “Mmdetection: Open mmlab detection toolbox and benchmark,” arXiv preprint arXiv:1906.07155, 2019.
[ 51 ] T.-Y. Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ramanan, P. Doll´ar, and C. L. Zitnick, “Microsoft coco: Common objects in context,” in European conference on computer vision, pp. 740–755, Springer, 2014.
[ 52 ] J. Hosang, R. Benenson, and B. Schiele, “Learning non-maximum suppression,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 4507–4515, 2017.
[ 53 ] P. Bilic, P. F. Christ, E. Vorontsov, G. Chlebus, H. Chen, Q. Dou, C.-W. Fu, X. Han, P.-A. Heng, J. Hesser, et al., “The liver tumor segmentation benchmark (lits),” arXiv preprint arXiv:1901.04056, 2019.
[ 54 ] L. Soler, A. Hostettler, V. Agnus, A. Charnoz, J. Fasquel, J. Moreau, A. Osswald, M. Bouhadjar, and J. Marescaux, “3d image reconstruction for comparison of algorithm database: A patient specific anatomical and medical image database,” IRCAD, Strasbourg, France, Tech. Rep, 2010.
[ 55 ] A. A. Taha and A. Hanbury, “Metrics for evaluating 3d medical image segmentation: analysis, selection, and tool,” BMC medical imaging, vol. 15, no. 1, pp. 1–28, 2015.
[ 56 ] Y. Tang, Y. Tang, Y. Zhu, J. Xiao, and R. M. Summers, “E^2net: An edge enhanced network for accurate liver and tumor segmentation on ct scans,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 512–522, Springer, 2020.
[ 57 ] X. Li, H. Chen, X. Qi, Q. Dou, C.-W. Fu, and P.-A. Heng, “H-denseunet: hybrid densely connected unet for liver and tumor segmentation from ct volumes,” IEEE transactions on medical imaging, vol. 37, no. 12, pp. 2663–2674, 2018.
[ 58 ] X. Chen, R. Zhang, and P. Yan, “Feature fusion encoder decoder network for automatic liver lesion segmentation,” in 2019 IEEE 16th international symposium on biomedical imaging (ISBI 2019), pp. 430–433, IEEE, 2019.
[ 59 ] L. Bi, J. Kim, A. Kumar, and D. Feng, “Automatic liver lesion detection using cascaded deep residual networks,” arXiv preprint arXiv:1704.02703, 2017.
[ 60 ] Y. Yuan, “Hierarchical convolutional-deconvolutional neural networks for automatic liver and tumor segmentation,” arXiv preprint arXiv:1710.04540, 2017.
[ 61 ] E. Vorontsov, A. Tang, C. Pal, and S. Kadoury, “Liver lesion segmentation informed by joint liver segmentation,” in 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), pp. 1332–1335, IEEE, 2018.
[ 62 ] G. Chlebus, A. Schenk, J. H. Moltz, B. van Ginneken, H. K. Hahn, and H. Meine, “Automatic liver tumor segmentation in ct with fully convolutional neural networks and object-based postprocessing,” Scientific reports, vol. 8, no. 1, pp. 1–7, 2018.
[ 63 ] J. Tian, C. Li, Z. Shi, and F. Xu, “A diagnostic report generator from ct volumes on liver tumor with semi-supervised attention mechanism,” in International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 702–710, Springer, 2018.
[ 64 ] R. Dey and Y. Hong, “Hybrid cascaded neural network for liver lesion segmentation,” in 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI), pp. 1173–1177, IEEE, 2020.
[ 65 ] E. Vorontsov, M. Cerny, P. R´egnier, L. Di Jorio, C. J. Pal, R. Lapointe, F. Vandenbroucke-Menu, S. Turcotte, S. Kadoury, and A. Tang, “Deep learning for automated segmentation of liver lesions at ct in patients with colorectal cancer liver metastases,” Radiology: Artificial Intelligence, vol. 1, no. 2, p. 180014, 2019.
[ 66 ] J. G. Nam, S. Park, E. J. Hwang, J. H. Lee, K.-N. Jin, K. Y. Lim, T. H. Vu, J. H. Sohn, S. Hwang, J. M. Goo, et al., “Development and validation of deep learning–based automatic detection algorithm for malignant pulmonary nodules on chest radiographs,” Radiology, vol. 290, no. 1, pp. 218–228, 2019.
[ 67 ] K. Chen, J. Pang, J. Wang, Y. Xiong, X. Li, S. Sun, W. Feng, Z. Liu, J. Shi, W. Ouyang, et al., “Hybrid task cascade for instance segmentation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4974–4983, 2019.
[ 68 ] S. Qiao, L.-C. Chen, and A. Yuille, “Detectors: Detecting objects with recursive feature pyramid and switchable atrous convolution,” arXiv preprint arXiv:2006.02334, 2020.