Segmentation-Free Estimation of Kidney Volumes in CT with Dual Regression Forests

详细信息    查看全文

• 刊名：Lecture Notes in Computer Science
• 出版年：2016
• 出版时间：2016
• 年：2016
• 卷：10019
• 期：1
• 页码：156-163
• 全文大小：1,561 KB
• 参考文献：1.Zelmer, J.L.: The economic burden of end-stage renal disease in canada. Kidney Int. Masson SAS. 72(9), 1122–1129 (2007)CrossRef
  2.Diez, A., Powelson, J., et al.: Correlation between CT-based measured renal volumes and nuclear-renography-based SRF in living kidney donors. clinical diagnostic utility and practice patterns. Clin. Transplant. 28(6), 675–682 (2014)CrossRef
  3.Connolly, J.O., Woolfson, R.G.: A critique of clinical guidelines for detection of individuals with chronic kidney disease. Neph. Clin. Pract. 111(1), c69–c73 (2009)CrossRef
  4.Widjaja, E., Oxtoby, J.W., et al.: Ultrasound measured renal length vs. low dose CT volume in predicting single kidney GFR. Br. J. Radiol. 77, 759–764 (2014)CrossRef
  5.Yan, G., Wang, B.: An automatic kidney segmentation from abdominal CT images. IEEE Intell. Comput. Intell. Syst. 1, 280–284 (2010)
  6.Li, X., Chen, X., Yao, J., Zhang, X., Tian, J.: Renal cortex segmentation using optimal surface search with novel graph construction. In: Fichtinger, G., Martel, A., Peters, T. (eds.) MICCAI 2011. LNCS, vol. 6893, pp. 387–394. Springer, Heidelberg (2011). doi:10.​1007/​978-3-642-23626-6_​48 CrossRef
  7.Zhang, Y., Matuszewski, B.J., Shark, L.K., Moore, C.J.: Medical image segmentation using new hybrid level-set method. In: BioMedical Visualization, pp. 71–76 (2008)
  8.Zhen, X., Wang, Z., Islam, A., Bhaduri, M., Chan, I., Li, S.: Multi-scale deep networks and regression forests for direct bi-ventricular volume estimation. Med. Image Anal. 30, 120–129 (2016)CrossRef
  9.Lin, D.T., Lei, C.C., Hung, S.W.: Computer-aided kidney segmentation on abdominal CT images. IEEE Trans. Inf. Tech. Biomed. 10(1), 59–65 (2006)CrossRef
  10.Cuingnet, R., Prevost, R., Lesage, D., Cohen, L.D., Mory, B., Ardon, R.: Automatic detection and segmentation of kidneys in 3D CT images using random forests. In: Ayache, N., Delingette, H., Golland, P., Mori, K. (eds.) MICCAI 2012. LNCS, vol. 7512, pp. 66–74. Springer, Heidelberg (2012). doi:10.​1007/​978-3-642-33454-2_​9 CrossRef
  11.Glocker, B., Pauly, O., Konukoglu, E., Criminisi, A.: Joint classification-regression forests for spatially structured multi-object segmentation. In: European Conference on Computer Vision, pp. 870–881 (2012)
  12.Thong, W., Kadoury, S., Piche, N., Pal, C.J.: Convolutional networks for kidney segmentation in contrast-enhanced CT scans. Comput. Methods Biomech. Biomed. Eng.: Imaging Visual., 1–6 (2016)
  13.Zhen, X., Wang, Z., Islam, A., Bhaduri, M., Chan, I., Li, S.: Direct estimation of cardiac bi-ventricular volumes with regression forests. In: Golland, P., Hata, N., Barillot, C., Hornegger, J., Howe, R. (eds.) MICCAI 2014. LNCS, vol. 8674, pp. 586–593. Springer, Heidelberg (2014). doi:10.​1007/​978-3-319-10470-6_​73
  
• 作者单位：Mohammad Arafat Hussain (18)
  Ghassan Hamarneh (19)
  Timothy W. O’Connell (20)
  Mohammed F. Mohammed (20)
  Rafeef Abugharbieh (18)
  
  18. Department of Electrical and Computer Engineering, University of British Columbia, Vancouver, BC, Canada
  19. School of Computing Science, Simon Fraser University, Burnaby, BC, Canada
  20. Division of Emergency and Trauma Radiology, Vancouver General Hospital, Vancouver, BC, Canada
  
• 丛书名：Machine Learning in Medical Imaging
• ISBN：978-3-319-47157-0
• 刊物类别：Computer Science
• 刊物主题：Artificial Intelligence and Robotics
  Computer Communication Networks
  Software Engineering
  Data Encryption
  Database Management
  Computation by Abstract Devices
  Algorithm Analysis and Problem Complexity
  
• 出版者：Springer Berlin / Heidelberg
• ISSN：1611-3349
• 卷排序：10019

文摘

Accurate estimation of kidney volume is essential for clinical diagnoses and therapeutic decisions related to renal diseases. Existing kidney volume estimation methods rely on an intermediate segmentation step that is subject to various limitations. In this work, we propose a segmentation-free, supervised learning approach that addresses the challenges of accurate kidney volume estimation caused by extensive variations in kidney shape, size and orientation across subjects. We develop dual regression forests to simultaneously predict the kidney area per image slice, and kidney span per image volume. We validate our method on a dataset of 45 subjects with a total of 90 kidney samples. We obtained a volume estimation accuracy higher than existing segmentation-free (by 72 %) and segmentation-based methods (by 82 %). Compared to a single regression model, the dual regression reduced the false positive area-estimates and improved volume estimation accuracy by 41 %. We also found a mean deviation of under 10 % between our estimated kidney volumes and those obtained manually by expert radiologists.