1 / 19

Automated Segmentation of Computed Tomography Images

Automated Segmentation of Computed Tomography Images. Justin Senseney Paul Hemler, PhD Matthew J. McAuliffe, PhD. Introduction. Chronic osteoarthritis risk factors, over time Obesity (NIH: Body Mass Index > 30) BMI discussion Obesity well measured by relative loss of muscle

seven
Download Presentation

Automated Segmentation of Computed Tomography Images

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Automated Segmentation of Computed Tomography Images Justin Senseney Paul Hemler, PhD Matthew J. McAuliffe, PhD

  2. Introduction • Chronic osteoarthritis risk factors, over time • Obesity (NIH: Body Mass Index > 30) • BMI discussion • Obesity well measured by relative loss of muscle • Fat around muscle tissue independent of BMI • Need for automated systems to quantitatively measure muscle loss K. F. Adams, A. Schatzkin, T. B. Harris, V. Kipnis, T. Mouw, R. Ballard-Barbash, A. Hollenbeck, and M. F. Leitzmann. Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old. New England Journal of Medicine, 355(8):763–778, 2006.

  3. Methods • Automatic segmentation • Semiautomatic segmentation M. McAuliffe, F. Lalonde, D. McGarry, W. Gandler, K. Csaky, and B. Trus. Medical image processing, analysis and visualization in clinical research. In Computer- Based Medical Systems, 2001. CBMS 2001. Proceedings. 14th IEEE Symposium on, pages 381–386, 2001.

  4. Automatic Segmentation – Thigh • Thigh segmentation • Threshold • Connected thigh segmentation • Identify separation S. Ohshima, S. Yamamoto, T. Yamaji, M. Suzuki, M. Mutoh, M. Iwasaki, S. Sasazuki, K. Kotera, S. Tsugane, Y. Muramatsu, and N. Moriyama. Development of an automated 3d segmentation program for volume quantification of body fat distribution using ct. Japanese Journal of Radiological Technology, 64(9):1177–1181, 2008.

  5. Automatic Segmentation – Thigh (2) • Bone segmentation • Region growing • Bone scattering • Marrow segmentation • Fascia, muscle segmentation

  6. Automatic Segmentation - Abdomen • Abdomen segmentation • Subcutaneous fat segmentation • Method from Zhao, et al. B. Zhao, J. Colville, J. Kalaigian, S. Curren, J. Li, P. Kijewski, and L. Schwartz. Automated quantification of body fat distribution on volumetric computed tomography. Journal of Computer Assisted Tomography, 30(5), 2006.

  7. Semi-Automatic Segmentation • 2D options: • Livewire • Level set • 2D/3D options: • Region growing • B-spline approximations during slice propagation

  8. Tissue Classification • Partial voluming concerns: • -190 <= fat pixel <= -30 • 0 <= muscle pixel <= 100 • -30 < partial volume pixel < 0 • Custom look-up table: fat-red, muscle-blue, partial volume-white. S. Ohshima, S. Yamamoto, T. Yamaji, M. Suzuki, M. Mutoh, M. Iwasaki, S. Sasazuki, K. Kotera, S. Tsugane, Y. Muramatsu, and N. Moriyama. Development of an automated 3d segmentation program for volume quantification of body fat distribution using ct. Japanese Journal of Radiological Technology, 64(9):1177–1181, 2008.

  9. Muscle and Fat Quantification • Reports • Text • PDF, using iText • Standard output • MIPAV Statistics generator B. Lowagie. iText in Action: Creating and Manipulating PDF. Manning, New York, 2006.

  10. Customization • Interface for customized CT projects, options: • New regions of interest • Calculation dependencies • Display options • Calculation options Start Pane: Abdomen Start Voi: Abdomen Color: 255,200,0 Do_Calc: true End Voi Start Voi: Subcut. Color: 255,0,0 Do_Calc: true End Voi Start Voi: Phantom Color: 0,255,0 End Voi End Pane Start Pane: Tissue Start Voi: Visceral Color: 255,200,0 Do_Calc: true End Voi Start Voi: Liver Color: 255,0,0 End Voi Start Voi: Liver cysts Num_Curves: 7 Color: 0,255,0 Do_Calc: true Do_Fill: true End Voi Start Voi: Bone sample Color: 0,255,255 End Voi Start Voi: Water sample Color: 255,0,255 End Voi End Pane

  11. Customization Options • Orientation invariant • Volume/Area Options • Units Specification

  12. Thigh Results • Compared to 13 freehand segmented images from the University of California, San Diego (UCSD) • Useful for manually demanding segmentations

  13. Abdomen Results • Larger variability • Needs manual attention

  14. Conclusion • Useful automatic and semi-automatic methods • Ability to later refine these • Benefits from manual overview • Larger analysis set needed • Comparison to automatic methods • Comparison to other qualified people for manual segmentation

  15. Download • http://mipav.cit.nih.gov • Look in the plugins folder for the MuscleSegmentation plugin • SenseneyJ@mail.nih.gov • Open Source? No….

  16. Acknowledgments This work was supported by the Intramural Research Program of the National Institutes of Health and the Center for Information Technology at the National Institutes of Health.

  17. Questions?

  18. Watershed? • Requires pre-processing steps • Limited viability

  19. Future work? • Algorithms • Data • Usability • Extensibility

More Related