Korean Journal of Radiology

The Korean Society of Radiology (대한영상의학회)
권호 표지
DOI QR코드

DOI QR Code

Deep Learning Algorithm for Automated Segmentation and Volume Measurement of the Liver and Spleen Using Portal Venous Phase Computed Tomography Images

  • Yura Ahn (Department of Radiology and Research Institute of Radiology, Asan Medical Center, University of Ulsan College of Medicine) ;
  • Jee Seok Yoon (Department of Brain and Cognitive Engineering, Korea University) ;
  • Seung Soo Lee (Department of Radiology and Research Institute of Radiology, Asan Medical Center, University of Ulsan College of Medicine) ;
  • Heung-Il Suk (Department of Brain and Cognitive Engineering, Korea University) ;
  • Jung Hee Son (Department of Radiology and Research Institute of Radiology, Asan Medical Center, University of Ulsan College of Medicine) ;
  • Yu Sub Sung (Department of Radiology and Research Institute of Radiology, Asan Medical Center, University of Ulsan College of Medicine) ;
  • Yedaun Lee (Department of Radiology, Haeundae Paik Hospital, Inje University College of Medicine) ;
  • Bo-Kyeong Kang (Department of Radiology, Hanyang University Medical Center, Hanyang University School of Medicine) ;
  • Ho Sung Kim (Department of Radiology and Research Institute of Radiology, Asan Medical Center, University of Ulsan College of Medicine)
  • Received : 2020.03.06
  • Accepted : 2020.05.11
  • Published : 2020.08.01
Download PDF
⟨ Previous Next ⟩

Abstract

Objective: Measurement of the liver and spleen volumes has clinical implications. Although computed tomography (CT) volumetry is considered to be the most reliable noninvasive method for liver and spleen volume measurement, it has limited application in clinical practice due to its time-consuming segmentation process. We aimed to develop and validate a deep learning algorithm (DLA) for fully automated liver and spleen segmentation using portal venous phase CT images in various liver conditions. Materials and Methods: A DLA for liver and spleen segmentation was trained using a development dataset of portal venous CT images from 813 patients. Performance of the DLA was evaluated in two separate test datasets: dataset-1 which included 150 CT examinations in patients with various liver conditions (i.e., healthy liver, fatty liver, chronic liver disease, cirrhosis, and post-hepatectomy) and dataset-2 which included 50 pairs of CT examinations performed at ours and other institutions. The performance of the DLA was evaluated using the dice similarity score (DSS) for segmentation and Bland-Altman 95% limits of agreement (LOA) for measurement of the volumetric indices, which was compared with that of ground truth manual segmentation. Results: In test dataset-1, the DLA achieved a mean DSS of 0.973 and 0.974 for liver and spleen segmentation, respectively, with no significant difference in DSS across different liver conditions (p = 0.60 and 0.26 for the liver and spleen, respectively). For the measurement of volumetric indices, the Bland-Altman 95% LOA was -0.17 ± 3.07% for liver volume and -0.56 ± 3.78% for spleen volume. In test dataset-2, DLA performance using CT images obtained at outside institutions and our institution was comparable for liver (DSS, 0.982 vs. 0.983; p = 0.28) and spleen (DSS, 0.969 vs. 0.968; p = 0.41) segmentation. Conclusion: The DLA enabled highly accurate segmentation and volume measurement of the liver and spleen using portal venous phase CT images of patients with various liver conditions.

Keywords

Acknowledgement

Seung Soo Lee was responsible for the clinical study, and Heung-Il Suk was responsible for the development of the deep learning algorithm. Both authors contributed equally to this study and assume the role of corresponding authors.

References

  1. Lim MC, Tan CH, Cai J, Zheng J, Kow AW. CT volumetry of the liver: where does it stand in clinical practice? Clin Radiol 2014;69:887-895 https://doi.org/10.1016/j.crad.2013.12.021
  2. Schindl MJ, Redhead DN, Fearon KC, Garden OJ, Wigmore SJ; Edinburgh Liver Surgery and Transplantation Experimental Research Group (eLISTER). The value of residual liver volume as a predictor of hepatic dysfunction and infection after major liver resection. Gut 2005;54:289-296 https://doi.org/10.1136/gut.2004.046524
  3. Ogasawara K, Une Y, Nakajima Y, Uchino J. The significance of measuring liver volume using computed tomographic images before and after hepatectomy. Surgery Today 1995;25:43-48 https://doi.org/10.1007/BF00309384
  4. Prodeau M, Drumez E, Duhamel A, Vibert E, Farges O, Lassailly G, et al. An ordinal model to predict the risk of symptomatic liver failure in patients with cirrhosis undergoing hepatectomy. J Hepatol 2019;71:920-929 https://doi.org/10.1016/j.jhep.2019.06.003
  5. Huang Y, Huang B, Kan T, Yang B, Yuan M, Wang J. Liver-tospleen ratio as an index of chronic liver diseases and safety of hepatectomy: a pilot study. World J Surg 2014;38:3186-3192 https://doi.org/10.1007/s00268-014-2717-6
  6. Berzigotti A, Seijo S, Arena U, Abraldes JG, Vizzutti F, GarciaPagan JC, et al. Elastography, spleen size, and platelet count identify portal hypertension in patients with compensated cirrhosis. Gastroenterology 2013;144:102-111.e1 https://doi.org/10.1053/j.gastro.2012.10.001
  7. Iranmanesh P, Vazquez O, Terraz S, Majno P, Spahr L, Poncet A, et al. Accurate computed tomography-based portal pressure assessment in patients with hepatocellular carcinoma. J Hepatol 2014;60:969-974 https://doi.org/10.1016/j.jhep.2013.12.015
  8. Murata Y, Abe M, Hiasa Y, Azemoto N, Kumagi T, Furukawa S, et al. Liver/spleen volume ratio as a predictor of prognosis in primary biliary cirrhosis. J Gastroenterol 2008;43:632-636 https://doi.org/10.1007/s00535-008-2202-9
  9. Pickhardt PJ, Malecki K, Hunt OF, Beaumont C, Kloke J, Ziemlewicz TJ, et al. Hepatosplenic volumetric assessment at MDCT for staging liver fibrosis. Eur Radiol 2017;27:3060-3068 https://doi.org/10.1007/s00330-016-4648-0
  10. Son JH, Lee SS, Lee Y, Kang BK, Sung YS, Jo S, et al. Assessment of liver fibrosis severity using computed tomography-based liver and spleen volumetric indices in patients with chronic liver disease. Eur Radiol 2020;30:3486-3496 https://doi.org/10.1007/s00330-020-06665-4
  11. Gotra A, Sivakumaran L, Chartrand G, Vu KN, VandenbrouckeMenu F, Kauffmann C, et al. Liver segmentation: indications, techniques and future directions. Insights Imaging 2017;8:377-392 https://doi.org/10.1007/s13244-017-0558-1
  12. Hu P, Wu F, Peng J, Liang P, Kong D. Automatic 3D liver segmentation based on deep learning and globally optimized surface evolution. Phys Med Biol 2016;61:8676-8698 https://doi.org/10.1088/1361-6560/61/24/8676
  13. Nakayama Y, Li Q, Katsuragawa S, Ikeda R, Hiai Y, Awai K, et al. Automated hepatic volumetry for living related liver transplantation at multisection CT. Radiology 2006;240:743-748 https://doi.org/10.1148/radiol.2403050850
  14. Fananapazir G, Bashir MR, Marin D, Boll DT. Computeraided liver volumetry: performance of a fully-automated, prototype post-processing solution for whole-organ and lobar segmentation based on MDCT imaging. Abdom Imaging 2015;40:1203-1212 https://doi.org/10.1007/s00261-014-0276-9
  15. Huynh HT, Karademir I, Oto A, Suzuki K. Computerized liver volumetry on MRI by using 3D geodesic active contour segmentation. AJR Am J Roentgenol 2014;202:152-159 https://doi.org/10.2214/AJR.13.10812
  16. Grieser C, Denecke T, Rothe JH, Geisel D, Stelter L, Cannon Walter T, et al. Gd-EOB enhanced MRI T1-weighted 3D-GRE with and without elevated flip angle modulation for threshold-based liver volume segmentation. Acta Radiol 2015;56:1419-1427 https://doi.org/10.1177/0284185114558975
  17. Huo Y, Terry JG, Wang J, Nair S, Lasko TA, Freedman BI, et al. Fully automatic liver attenuation estimation combing CNN segmentation and morphological operations. Med Phys 2019;46:3508-3519 https://doi.org/10.1002/mp.13675
  18. Wang K, Mamidipalli A, Retson T, Bahrami N, Hasenstab K, Blansit K, et al. Automated CT and MRI liver segmentation and biometry using a generalized convolutional neural network. Radiology: Artificial Intelligence 2019;1:e180022
  19. Choi KJ, Jang JK, Lee SS, Sung YS, Shim WH, Kim HS, et al. Development and validation of a deep learning system for staging liver fibrosis by using contrast agent-enhanced CT images in the Liver. Radiology 2018;289:688-697 https://doi.org/10.1148/radiol.2018180763
  20. Dou Q, Yu L, Chen H, Jin Y, Yang X, Qin J, et al. 3D deeply supervised network for automated segmentation of volumetric medical images. Med Image Anal 2017;41:40-54 https://doi.org/10.1016/j.media.2017.05.001
  21. Chen LC, Zhu Y, Papandreou G, Schroff F, Adam H. Encoderdecoder with atrous separable convolution for semantic image segmentation. The European conference on computer vision (ECCV);2018 September 8-14;Munich, Germany
  22. Roth HR, Lu L, Seff A, Cherry KM, Hoffman J, Wang S, et al. A new 2.5D representation for lymph node detection using random sets of deep convolutional neural network observations. Med Image Comput Comput Assist Interv 2014;17:520-527 https://doi.org/10.1007/978-3-319-10404-1_65
  23. Hashimoto T, Sugawara Y, Tamura S, Hasegawa K, Kishi Y, Kokudo N, et al. Estimation of standard liver volume in Japanese living liver donors. J Gastroenterol Hepatol 2006;21:1710-1713 https://doi.org/10.1111/j.1440-1746.2006.04433.x
  24. Serai SD, Obuchowski NA, Venkatesh SK, Sirlin CB, Miller FH, Ashton E, et al. Repeatability of MR elastography of liver: a meta-analysis. Radiology 2017;285:92-100 https://doi.org/10.1148/radiol.2017161398
  25. Saito A, Yamamoto S, Nawano S, Shimizu A. Automated liver segmentation from a postmortem CT scan based on a statistical shape model. Int J Comput Assist Radiol Surg 2017;12:205-221 https://doi.org/10.1007/s11548-016-1481-5
  26. Barnhart HX, Barboriak DP. Applications of the repeatability of quantitative imaging biomarkers: a review of statistical analysis of repeat data sets. Transl Oncol 2009;2:231-235 https://doi.org/10.1593/tlo.09268
  27. Prionas ND, Ray S, Boone JM. Volume assessment accuracy in computed tomography: a phantom study. J Appl Clin Med Phys 2010;11:3037
  28. Hori M, Suzuki K, Epstein ML, Baron RL. Computed tomography liver volumetry using 3-dimensional image data in living donor liver transplantation: effects of the slice thickness on the volume calculation. Liver Transpl 2011;17:1427-1436 https://doi.org/10.1002/lt.22419
  29. Gotra A, Chartrand G, Vu KN, Vandenbroucke-Menu F, Massicotte-Tisluck K, de Guise JA, et al. Comparison of MRIand CT-based semiautomated liver segmentation: a validation study. Abdom Radiol (NY) 2017;42:478-489 https://doi.org/10.1007/s00261-016-0912-7
  30. Suzuki K, Kohlbrenner R, Epstein ML, Obajuluwa AM, Xu J, Hori M. Computer-aided measurement of liver volumes in CT by means of geodesic active contour segmentation coupled with level-set algorithms. Med Phys 2010;37:2159-2166 https://doi.org/10.1118/1.3395579
  31. Harris A, Kamishima T, Hao HY, Kato F, Omatsu T, Onodera Y, et al. Splenic volume measurements on computed tomography utilizing automatically contouring software and its relationship with age, gender, and anthropometric parameters. Eur J Radiol 2010;75:e97-e101 https://doi.org/10.1016/j.ejrad.2009.08.013
  32. Dello SA, Stoot JH, van Stiphout RS, Bloemen JG, Wigmore SJ, Dejong CH, et al. Prospective volumetric assessment of the liver on a personal computer by nonradiologists prior to partial hepatectomy. World J Surg 2011;35:386-392 https://doi.org/10.1007/s00268-010-0877-6