Abstract

Automatic tongue area segmentation is crucial for computer aided tongue diagnosis, but traditional intensity-based segmentation methods that make use of monochromatic images cannot provide accurate and robust results. We propose a novel tongue segmentation method that uses hyperspectral images and the support vector machine. This method combines spatial and spectral information to analyze the medical tongue image and can provide much better tongue segmentation results. The promising experimental results and quantitative evaluations demonstrate that our method can provide much better performance than the traditional method.

© 2007 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |

  1. T. Vo-Dinh, ed., Biomedical Photonics Handbook (CRC, 2003).
    [CrossRef]
  2. R. B. Bell, D. Kademani, L. Homer, E. J. Dierks, and B. E. Potter, "Tongue cancer: is there a difference in survival compared with other subsites in the oral cavity?"J. Oral Maxillofac. Surg. 65, 229-236 (2007).
    [CrossRef] [PubMed]
  3. J. A. Byrd, A. J. Bruce, and R. S. Rogers, "Glossitis and other tongue disorders," Dermatol. Clin. 21, 123-134 (2003).
    [CrossRef] [PubMed]
  4. B. Pang, D. Zhang, and K. Q. Wang, "The bi-elliptical deformable contour and its application to automated tongue segmentation in Chinese medicine," IEEE Trans. Med. Imaging 24, 946-956 (2005).
    [CrossRef] [PubMed]
  5. I. Feng and S. Tianbin, Practicality handbook of tongue diagnosis of TC (in Chinese) (Ke Xue Chu Ban She, 2002).
  6. W. Zuo, K. Wang, D. Zhang, and H. Zhang, "Combination of polar edge detection and active contour model for automated tongue segmentation," in Proceedings of Third International Conference on Image and Graphics (IEEE, 2004), pp. 270-273.
  7. T. Vo-Dinh, "A hyperspectral imaging system for in vivo optical diagnostics," IEEE Eng. Med. Biol. Mag. 23, 40-49 (2004).
    [CrossRef] [PubMed]
  8. N. Magotra, E. Wu, P. Soliz, P. Truitt, P. Gelabert, and T. Stetzler, "Hyperspectral biomedical image formation," Thirty-Third Asilomar Conference on Signals, Systems, and Computers (IEEE, 1999), pp. 462-465.
    [CrossRef]
  9. V. Tuchin, Tissue optics: light scattering methods and instruments for medical diagnosis (SPIE Press, 2000).
  10. R. Anderson and J. Parrish, "The Optics of Human Skin," J. Invest. Dermatol. 77, pp. 13-19 (1981).
    [CrossRef]
  11. E. Edwards and S. Duntley, "The pigments and color of living human skin," Am. J. Anat. 65, pp. 1-33 (1939).
    [CrossRef]
  12. E. Angelopoulou, R. Molana, and K. Daniilidis, "Multispectral skin color modeling," in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (IEEE, 2001), pp. 635-642.
  13. Z. Pan and G. Healey, "Face recognition in hyperspectral images," IEEE Trans. Pattern Anal. Mach. Intell. 25, pp. 1552-1560 (2003).
  14. W. R. Crum, T. Hartkens, and D. L. G. Hill, "Non-rigid image registration: theory and practice," Br. J. Radiol. 77, pp. S140-S153 (2004).
    [CrossRef]
  15. F. Melgani and L. Bruzzone, "Classification of hyperspectral remotesensing images with support vector machines," IEEE Trans. Geosci. Remote Sens. 42, pp. 1778-1790 (2004).
    [CrossRef]
  16. H. Bischof and A. Leona, "Finding optimal neural networks for land use classification," IEEE Trans. Geosci. Remote Sens. 36, pp. 337-341 (1998).
    [CrossRef]
  17. L. Bruzzone and D. Fernández-Prieto, "A technique for the selection of kernel-function parameters in RBF neural networks for classification of remote-sensing images," IEEE Trans. Geosci. Remote Sens. 37, pp. 1179-1184 (1999).
    [CrossRef]
  18. G. E. Hughes, "On the mean accuracy of statistical pattern recognizers," IEEE Trans. Inf. Theory 14, pp. 55-63 (1968).
    [CrossRef]
  19. P. Mouroulis, R. O. Green, and T. G. Chrien, "Pushbroom imaging spectrometer design for optimum recovery of spectroscopic and spatial information," Appl. Opt. 39, pp. 2210-2220 (2000).
    [CrossRef]
  20. M. B. Sinclair, J. A. Timlin, D. M. Haaland, and M. Werner-Washburne, "Design, construction, characterization, and application of a hyperspectral microarray scanner," Appl. Opt. 43, pp. 2079-2088 (2004).
  21. C. A. Shah, P. Watanachaturaporn, P. K. Varshney, and M. K. Arora, "Some recent results on hyperspectral image classification," in Proceedings of IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data (IEEE, 2003), pp. 346-353.
    [CrossRef]
  22. V. N. Vapnik, The Nature of Statistical Learning Theory (Springer-Verlag, 1995).
  23. G. Camps-Valls and L. Bruzzone, "Kernel-based methods for hyperspectral image classification," IEEE Trans. Geosci. Remote Sens. 43, pp. 1351-1362 (2005).
    [CrossRef]
  24. V. Chalana and Y. Kim, "A methodology for evaluation of boundary detection algorithms on medical images," IEEE Trans. Med. Imaging 16, pp. 642-652 (1997).
    [CrossRef]

2007 (1)

R. B. Bell, D. Kademani, L. Homer, E. J. Dierks, and B. E. Potter, "Tongue cancer: is there a difference in survival compared with other subsites in the oral cavity?"J. Oral Maxillofac. Surg. 65, 229-236 (2007).
[CrossRef] [PubMed]

2005 (2)

B. Pang, D. Zhang, and K. Q. Wang, "The bi-elliptical deformable contour and its application to automated tongue segmentation in Chinese medicine," IEEE Trans. Med. Imaging 24, 946-956 (2005).
[CrossRef] [PubMed]

G. Camps-Valls and L. Bruzzone, "Kernel-based methods for hyperspectral image classification," IEEE Trans. Geosci. Remote Sens. 43, pp. 1351-1362 (2005).
[CrossRef]

2004 (3)

T. Vo-Dinh, "A hyperspectral imaging system for in vivo optical diagnostics," IEEE Eng. Med. Biol. Mag. 23, 40-49 (2004).
[CrossRef] [PubMed]

W. R. Crum, T. Hartkens, and D. L. G. Hill, "Non-rigid image registration: theory and practice," Br. J. Radiol. 77, pp. S140-S153 (2004).
[CrossRef]

F. Melgani and L. Bruzzone, "Classification of hyperspectral remotesensing images with support vector machines," IEEE Trans. Geosci. Remote Sens. 42, pp. 1778-1790 (2004).
[CrossRef]

2003 (2)

J. A. Byrd, A. J. Bruce, and R. S. Rogers, "Glossitis and other tongue disorders," Dermatol. Clin. 21, 123-134 (2003).
[CrossRef] [PubMed]

Z. Pan and G. Healey, "Face recognition in hyperspectral images," IEEE Trans. Pattern Anal. Mach. Intell. 25, pp. 1552-1560 (2003).

2000 (1)

P. Mouroulis, R. O. Green, and T. G. Chrien, "Pushbroom imaging spectrometer design for optimum recovery of spectroscopic and spatial information," Appl. Opt. 39, pp. 2210-2220 (2000).
[CrossRef]

1999 (1)

L. Bruzzone and D. Fernández-Prieto, "A technique for the selection of kernel-function parameters in RBF neural networks for classification of remote-sensing images," IEEE Trans. Geosci. Remote Sens. 37, pp. 1179-1184 (1999).
[CrossRef]

1998 (1)

H. Bischof and A. Leona, "Finding optimal neural networks for land use classification," IEEE Trans. Geosci. Remote Sens. 36, pp. 337-341 (1998).
[CrossRef]

1997 (1)

V. Chalana and Y. Kim, "A methodology for evaluation of boundary detection algorithms on medical images," IEEE Trans. Med. Imaging 16, pp. 642-652 (1997).
[CrossRef]

1981 (1)

R. Anderson and J. Parrish, "The Optics of Human Skin," J. Invest. Dermatol. 77, pp. 13-19 (1981).
[CrossRef]

1968 (1)

G. E. Hughes, "On the mean accuracy of statistical pattern recognizers," IEEE Trans. Inf. Theory 14, pp. 55-63 (1968).
[CrossRef]

1939 (1)

E. Edwards and S. Duntley, "The pigments and color of living human skin," Am. J. Anat. 65, pp. 1-33 (1939).
[CrossRef]

Am. J. Anat. (1)

E. Edwards and S. Duntley, "The pigments and color of living human skin," Am. J. Anat. 65, pp. 1-33 (1939).
[CrossRef]

Appl. Opt. (2)

P. Mouroulis, R. O. Green, and T. G. Chrien, "Pushbroom imaging spectrometer design for optimum recovery of spectroscopic and spatial information," Appl. Opt. 39, pp. 2210-2220 (2000).
[CrossRef]

M. B. Sinclair, J. A. Timlin, D. M. Haaland, and M. Werner-Washburne, "Design, construction, characterization, and application of a hyperspectral microarray scanner," Appl. Opt. 43, pp. 2079-2088 (2004).

Br. J. Radiol. (1)

W. R. Crum, T. Hartkens, and D. L. G. Hill, "Non-rigid image registration: theory and practice," Br. J. Radiol. 77, pp. S140-S153 (2004).
[CrossRef]

Dermatol. Clin. (1)

J. A. Byrd, A. J. Bruce, and R. S. Rogers, "Glossitis and other tongue disorders," Dermatol. Clin. 21, 123-134 (2003).
[CrossRef] [PubMed]

IEEE Eng. Med. Biol. Mag. (1)

T. Vo-Dinh, "A hyperspectral imaging system for in vivo optical diagnostics," IEEE Eng. Med. Biol. Mag. 23, 40-49 (2004).
[CrossRef] [PubMed]

IEEE Trans. Geosci. Remote Sens. (4)

F. Melgani and L. Bruzzone, "Classification of hyperspectral remotesensing images with support vector machines," IEEE Trans. Geosci. Remote Sens. 42, pp. 1778-1790 (2004).
[CrossRef]

H. Bischof and A. Leona, "Finding optimal neural networks for land use classification," IEEE Trans. Geosci. Remote Sens. 36, pp. 337-341 (1998).
[CrossRef]

L. Bruzzone and D. Fernández-Prieto, "A technique for the selection of kernel-function parameters in RBF neural networks for classification of remote-sensing images," IEEE Trans. Geosci. Remote Sens. 37, pp. 1179-1184 (1999).
[CrossRef]

G. Camps-Valls and L. Bruzzone, "Kernel-based methods for hyperspectral image classification," IEEE Trans. Geosci. Remote Sens. 43, pp. 1351-1362 (2005).
[CrossRef]

IEEE Trans. Inf. Theory (1)

G. E. Hughes, "On the mean accuracy of statistical pattern recognizers," IEEE Trans. Inf. Theory 14, pp. 55-63 (1968).
[CrossRef]

IEEE Trans. Med. Imaging (2)

B. Pang, D. Zhang, and K. Q. Wang, "The bi-elliptical deformable contour and its application to automated tongue segmentation in Chinese medicine," IEEE Trans. Med. Imaging 24, 946-956 (2005).
[CrossRef] [PubMed]

V. Chalana and Y. Kim, "A methodology for evaluation of boundary detection algorithms on medical images," IEEE Trans. Med. Imaging 16, pp. 642-652 (1997).
[CrossRef]

IEEE Trans. Pattern Anal. Mach. Intell. (1)

Z. Pan and G. Healey, "Face recognition in hyperspectral images," IEEE Trans. Pattern Anal. Mach. Intell. 25, pp. 1552-1560 (2003).

J. Invest. Dermatol. (1)

R. Anderson and J. Parrish, "The Optics of Human Skin," J. Invest. Dermatol. 77, pp. 13-19 (1981).
[CrossRef]

J. Oral Maxillofac. Surg. (1)

R. B. Bell, D. Kademani, L. Homer, E. J. Dierks, and B. E. Potter, "Tongue cancer: is there a difference in survival compared with other subsites in the oral cavity?"J. Oral Maxillofac. Surg. 65, 229-236 (2007).
[CrossRef] [PubMed]

Other (8)

T. Vo-Dinh, ed., Biomedical Photonics Handbook (CRC, 2003).
[CrossRef]

N. Magotra, E. Wu, P. Soliz, P. Truitt, P. Gelabert, and T. Stetzler, "Hyperspectral biomedical image formation," Thirty-Third Asilomar Conference on Signals, Systems, and Computers (IEEE, 1999), pp. 462-465.
[CrossRef]

V. Tuchin, Tissue optics: light scattering methods and instruments for medical diagnosis (SPIE Press, 2000).

I. Feng and S. Tianbin, Practicality handbook of tongue diagnosis of TC (in Chinese) (Ke Xue Chu Ban She, 2002).

W. Zuo, K. Wang, D. Zhang, and H. Zhang, "Combination of polar edge detection and active contour model for automated tongue segmentation," in Proceedings of Third International Conference on Image and Graphics (IEEE, 2004), pp. 270-273.

E. Angelopoulou, R. Molana, and K. Daniilidis, "Multispectral skin color modeling," in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (IEEE, 2001), pp. 635-642.

C. A. Shah, P. Watanachaturaporn, P. K. Varshney, and M. K. Arora, "Some recent results on hyperspectral image classification," in Proceedings of IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data (IEEE, 2003), pp. 346-353.
[CrossRef]

V. N. Vapnik, The Nature of Statistical Learning Theory (Springer-Verlag, 1995).

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (10)

Fig. 1
Fig. 1

Schematic of hyperspectral imaging sensor system.

Fig. 2
Fig. 2

(Color online) (a) HSI cube and (b) spectrum corresponding to the point in (a).

Fig. 3
Fig. 3

Flow chart of hyperspectral tongue image segmentation procedure.

Fig. 4
Fig. 4

ROI in the captured image (the left) and the binary image representing the detected tongue area (the right).

Fig. 5
Fig. 5

(Color online) (a) Spectrum of the visible light. (b) Top row: RGB tongue image synthesized by three spectral images. Bottom row: segmentation results in hyperspectral images and mapping them directly into the RGB tongue image.

Fig. 6
Fig. 6

Series of visible spectral tongue images from a hyperspectral image cube.

Fig. 7
Fig. 7

Spectral curves of the tongue and the skin of the face from two subjects.

Fig. 8
Fig. 8

Block diagram of the architecture for solving multiclass problems with SVMs.

Fig. 9
Fig. 9

Comparison of tongue segmentation results (a). (c) and (e) are the results of the BEDC method [4]. (b), (d), and (f) are the results of our proposed HSI+SVM method.

Fig. 10
Fig. 10

(Color online) Histogram of the comparison of our method and BEDC: (a) H D ¯ and M D ¯ and (b) S E ( H D ) and S E ( M D ) .

Tables (1)

Tables Icon

Table 1 Comparison of HSI+SVM Method and the BEDC Method [4] Tongue Segmentation Results in Terms of Boundary Metrics

Equations (12)

Equations on this page are rendered with MathJax. Learn more.

I ( x , y , λ i ) = L ( x , y , λ i ) S ( x , y , λ i ) R ( x , y , λ i ) + O ( x , y , λ i ) ,
I W ( x , y , λ i ) = L ( x , y , λ i ) S ( x , y , λ i ) R W ( x , y , λ i ) + O ( x , y , λ i ) ,
I B ( x , y , λ i ) = L ( x , y , λ i ) S ( x , y , λ i ) R B ( x , y , λ i ) + O ( x , y , λ i ) ,
L ( x , y , λ i ) S ( x , y , λ i ) = I W ( x , y , λ i ) I B ( x , y , λ i ) R W ( λ i ) R B ( λ i ) .
O ( x , y , λ i ) = I B ( x , y , λ i ) I W ( x , y , λ i ) I B ( x , y , λ i ) R W ( λ i ) R B ( λ i ) × R B ( λ i ) .
R ( x , y , λ i ) = I ( x , y , λ i ) I B ( x , y , λ i ) I W ( x , y , λ i ) I B ( x , y , λ i ) R W ( λ i ) + I W ( x , y , λ i ) I ( x , y , λ i ) I W ( x , y , λ i ) I B ( x , y , λ i ) R B ( λ i ) .
R ˜ = R / R ,
d ( a i , B ) = min j b j a i .
H D ( A , B ) = max ( max i { d ( a i , B ) } , max j { d ( b j , A ) } ) .
M D ( A , B ) = 1 m + n ( i d ( a i , B ) + j d ( b j , A ) ) .
H D ¯ = i = 1 N H D / N M D ¯ = i = 1 N M D / N ,
S E = s N ,   s = 1 N 1 i = 1 N ( x i x ¯ ) 2 ,

Metrics