Decision Forests with Spatio-Temporal Features for Graph-Based Tumor Segmentation in 4D Lung CT

  • Hamidreza Mirzaei
  • Lisa Tang
  • Rene Werner
  • Ghassan Hamarneh
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8184)

Abstract

We propose an automatic lung tumor segmentation in dynamic CT images that incorporates the novel use of tumor tissue deformations. In contrast to elastography imaging techniques for measuring tumor tissue properties, which require mechanical compression and thereby interrupt normal breathing, we completely avoid the use of any external physical forces. Instead, we calculate the tissue deformations during normal respiration using deformable registration. We investigate machine learning methods in order to discover the spatio-temporal dynamics that would help distinguish tumor from normal tissue deformation patterns and integrate this information into the segmentation process. Our method adapts an ensemble of decision trees combined with a 3D graph-based optimization that takes into account spatio-temporal consistency. The experimental results on patients with large tumors achieved an average F-measure accuracy of 0.79.

Keywords

Lung tumor image segmentation registration CT images imbalance data machine learning graph-based optimization 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Kanakatte, et al.: A Pilot Study of Automatic Lung Tumor Segmentation from Positron Emission Tomography Images using Standard Uptake Values. IEEE CIISP, 363–368 (2007)Google Scholar
  2. 2.
    Dizaji, et al.: Evaluation of ultrasound in tissue characterization. In: Proc. EMBEC 1999, vol. 37(1), pp. 666–667 (1999)Google Scholar
  3. 3.
    Sudhakar, et al.: MR Elastography of Liver Tumors: Preliminary Results. AJR, 2781 (2008)Google Scholar
  4. 4.
    Dizaji, et al.: Detection of Internal Displacement of Tissues in Ultrasound Images Using Image Registration Technique. In: Proc. IEEE Canadian Conference on Electrical & Computer Engineering, pp. 1145–1148 (2002)Google Scholar
  5. 5.
    Muthupillai, et al.: Magnetic resonance elastography. Nat. Med. 2, 601–603 (1996)CrossRefGoogle Scholar
  6. 6.
    Rey, et al.: Automatic Detection and Segmentation of Evolving Processes in 3D Medical Images: Application to Multiple Sclerosis. MIA 6(2), 163–179 (2002)Google Scholar
  7. 7.
    Zhou, Liu: Training cost-sensitive neural networks with methods addressing class imbalance problem. IEEE TKDE 18(1), 63–77 (2006)Google Scholar
  8. 8.
    Handels, et al.: 4D medical image computing and visualization of lung tumor mobility in spatio-temporal CT image data. International J. Medical Informatics 76, S433–S439 (2007)Google Scholar
  9. 9.
    Richberg, et al.: Diffeomorphic diffusion registration of lung CT images. In: Medical Image Analysis for the Clinic: A Grand Challenge, MICCAI, pp. 55–62 (2010)Google Scholar
  10. 10.
    Murphy, et al.: Evaluation of Registration Methods on Thoracic CT: The EMPIRE10 Challenge. IEEE TMI 30(11), 1901–1920 (2011)Google Scholar
  11. 11.
    van Rikxoort, et al.: Automatic lung segmentation from thoracic computed tomography scans using a hybrid approach with error detection. Medical Physics 36(7), 2934–2947 (2009)CrossRefGoogle Scholar
  12. 12.
    Vercauteren, et al.: Diffeomorphic demons: Efficient non-parametric image registration. Neuroimage 45(1), S61–S72 (2008)Google Scholar
  13. 13.
    Boykov, et al.: An Experimental Comparison of Min-Cut/Max-Flow Algorithms for Energy Minimization in Vision. IEEE TPAMI 26(9), 1124–1137 (2004)CrossRefGoogle Scholar
  14. 14.
    Plathow, et al.: Measurement of tumor diameter-dependent mobility of lung tumors by dynamic MRI. Radiotherapy and Oncology 73(3), 349–354 (2004)CrossRefGoogle Scholar
  15. 15.
    Cuingnet, et al.: Spatial and anatomical regularization of SVM for brain image analysis. In: NIPS, pp. 460–468 (2010)Google Scholar
  16. 16.
    Tu, et al.: Automated extraction of the cortical sulci based on a supervised learning approach. IEEE TMI 26(4), 541–552 (2007)Google Scholar
  17. 17.
    Gu, et al.: Automated delineation of lung tumors from CT images using a single click ensemble segmentation approach. Radiotherapy and Oncology 105(2), 167–173 (2012)CrossRefGoogle Scholar
  18. 18.
    Syeda-Mahmood, Yang: Characterizing normal and abnormal cardiac echo motion patterns. IEEE Computers in Cardiology, 725–728 (2006)Google Scholar
  19. 19.
    Martin, et al.: Building motion models of lung tumours from cone-beam CT for radiotherapy applications. Physics in Medicine and Biology 58(6), 1809 (2013)CrossRefGoogle Scholar
  20. 20.
    Britton, et al.: Assessment of gross tumor volume regression and motion changes during radiotherapy for non-small-cell lung cancer as measured by four-dimensional computed tomography. Int. J. Radiat. Oncol. Biol. Phys. 68, 1036–1046 (2007)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2013

Authors and Affiliations

  • Hamidreza Mirzaei
    • 1
  • Lisa Tang
    • 2
  • Rene Werner
    • 3
  • Ghassan Hamarneh
    • 2
  1. 1.Computational Vision Lab.Simon Fraser UniversityCanada
  2. 2.Medical Image Analysis Lab.Simon Fraser UniversityCanada
  3. 3.Department of Computational NeuroscienceUniversity Medical Center Hamburg-EppendorfGermany

Personalised recommendations