Abstract

A novel statistical model based on texture and shape for fully automatic intraretinal layer segmentation of normal retinal tomograms obtained by a commercial 800nm optical coherence tomography (OCT) system is developed. While existing algorithms often fail dramatically due to strong speckle noise, non-optimal imaging conditions, shadows and other artefacts, the novel algorithm’s accuracy only slowly deteriorates when progressively increasing segmentation task difficulty. Evaluation against a large set of manual segmentations shows unprecedented robustness, even in the presence of additional strong speckle noise, with dynamic range tested down to 12dB, enabling segmentation of almost all intraretinal layers in cases previously inaccessible to the existing algorithms. For the first time, an error measure is computed from a large, representative manually segmented data set (466 B-scans from 17 eyes, segmented twice by different operators) and compared to the automatic segmentation with a difference of only 2.6% against the inter-observer variability.

© 2010 OSA

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. W. Drexler, and J. G. Fujimoto, Optical Coherence Tomography: Technology and Applications (Springer, 2008).
  2. T. Fabritius, S. Makita, M. Miura, R. Myllylä, and Y. Yasuno, “Automated segmentation of the macula by optical coherence tomography,” Opt. Express 17(18), 15659–15669 (2009).
    [CrossRef] [PubMed]
  3. R. J. Zawadzki, S. S. Choi, S. M. Jones, S. S. Oliver, and J. S. Werner, “Adaptive optics-optical coherence tomography: optimizing visualization of microscopic retinal structures in three dimensions,” J. Opt. Soc. Am. A 24(5), 1373 (2007).
    [CrossRef]
  4. M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
    [CrossRef] [PubMed]
  5. D. Cabrera Fernández, H. M. Salinas, and C. A. Puliafito, “Automated detection of retinal layer structures on optical coherence tomography images,” Opt. Express 13(25), 10200–10216 (2005).
    [CrossRef] [PubMed]
  6. M. Mujat, R. Chan, B. Cense, B. Park, C. Joo, T. Akkin, T. Chen, and J. de Boer, “Retinal nerve fiber layer thickness map determined from optical coherence tomography images,” Opt. Express 13(23), 9480–9491 (2005).
    [CrossRef] [PubMed]
  7. D. Koozekanani, K. Boyer, and C. Roberts, “Retinal thickness measurements from optical coherence tomography using a Markov boundary model,” IEEE Trans. Med. Imaging 20(9), 900–916 (2001).
    [CrossRef] [PubMed]
  8. D. Tolliver, Y. Koutis, H. Ishikawa, J. S. Schuman, and G. L. Miller, “Unassisted Segmentation of Multiple Retinal Layers via Spectral Rounding,” in ARVO(2008).
  9. A. Mishra, A. Wong, K. Bizheva, and D. A. Clausi, “Intra-retinal layer segmentation in optical coherence tomography images,” Opt. Express 17(26), 23719–23728 (2009).
    [CrossRef]
  10. I. W. Selesnick, R. G. Baraniuk, and N. G. Kingsbury, “The Dual-Tree Complex Wavelet Transform,” IEEE Signal Process. Mag. 22(6), 123–151 (2005).
    [CrossRef]
  11. A. Mishra, A. Wong, D. A. Clausi, and P. W. Fieguth, “Quasi-random nonlinear scale space,” Pattern Recognit. Lett. In Press. (Corrected Proof).
  12. A. Wong, A. Mishra, K. Bizheva, and D. A. Clausi, “General Bayesian estimation for speckle noise reduction in optical coherence tomography retinal imagery,” Opt. Express 18(8), 8338–8352 (2010).
    [CrossRef] [PubMed]
  13. P. Thevenaz, and M. Unser, “A pyramid approach to sub-pixel image fusion based on mutual information,” in Image Processing, 1996. Proceedings., International Conference on(1996), p. 265.
  14. C. O. S. Sorzano, P. Thevenaz, and M. Unser, “Elastic registration of biological images using vector-spline regularization,” BIEEE Biomed. Eng. 52(4), 652–663 (2005).
    [CrossRef]
  15. A. K. Mishra, P. W. Fieguth, and D. A. Clausi, “Decoupled Active Contour (DAC) for Boundary Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence 99.
  16. T. F. Cootes, G. J. Edwards, and C. J. Taylor, “Active Appearance Models,” IEEE Trans. Pattern Anal. Mach. Intell. 23(6), 681–685 (2001).
    [CrossRef]
  17. M. Scholz, M. Fraunholz, and J. Selbig, “Nonlinear Principal Component Analysis: Neural Network Models and Applications,” in Principal Manifolds for Data Visualization and Dimension Reduction(2007), pp. 44–67.
  18. M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
    [CrossRef] [PubMed]
  19. A. A. Efros, and W. T. Freeman, “Image quilting for texture synthesis and transfer,” in Proceedings of the 28th annual conference on Computer graphics and interactive techniques(ACM, 2001), pp. 341–346.

2010

2009

2008

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

2007

2005

I. W. Selesnick, R. G. Baraniuk, and N. G. Kingsbury, “The Dual-Tree Complex Wavelet Transform,” IEEE Signal Process. Mag. 22(6), 123–151 (2005).
[CrossRef]

D. Cabrera Fernández, H. M. Salinas, and C. A. Puliafito, “Automated detection of retinal layer structures on optical coherence tomography images,” Opt. Express 13(25), 10200–10216 (2005).
[CrossRef] [PubMed]

M. Mujat, R. Chan, B. Cense, B. Park, C. Joo, T. Akkin, T. Chen, and J. de Boer, “Retinal nerve fiber layer thickness map determined from optical coherence tomography images,” Opt. Express 13(23), 9480–9491 (2005).
[CrossRef] [PubMed]

C. O. S. Sorzano, P. Thevenaz, and M. Unser, “Elastic registration of biological images using vector-spline regularization,” BIEEE Biomed. Eng. 52(4), 652–663 (2005).
[CrossRef]

M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
[CrossRef] [PubMed]

2001

T. F. Cootes, G. J. Edwards, and C. J. Taylor, “Active Appearance Models,” IEEE Trans. Pattern Anal. Mach. Intell. 23(6), 681–685 (2001).
[CrossRef]

D. Koozekanani, K. Boyer, and C. Roberts, “Retinal thickness measurements from optical coherence tomography using a Markov boundary model,” IEEE Trans. Med. Imaging 20(9), 900–916 (2001).
[CrossRef] [PubMed]

Abramoff, M. M. D.

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

Akkin, T.

Baraniuk, R. G.

I. W. Selesnick, R. G. Baraniuk, and N. G. Kingsbury, “The Dual-Tree Complex Wavelet Transform,” IEEE Signal Process. Mag. 22(6), 123–151 (2005).
[CrossRef]

Bizheva, K.

Boyer, K.

D. Koozekanani, K. Boyer, and C. Roberts, “Retinal thickness measurements from optical coherence tomography using a Markov boundary model,” IEEE Trans. Med. Imaging 20(9), 900–916 (2001).
[CrossRef] [PubMed]

Cabrera Fernández, D.

Cense, B.

Chan, R.

Chen, T.

Choi, S. S.

Clausi, D. A.

Cootes, T. F.

T. F. Cootes, G. J. Edwards, and C. J. Taylor, “Active Appearance Models,” IEEE Trans. Pattern Anal. Mach. Intell. 23(6), 681–685 (2001).
[CrossRef]

de Boer, J.

Edwards, G. J.

T. F. Cootes, G. J. Edwards, and C. J. Taylor, “Active Appearance Models,” IEEE Trans. Pattern Anal. Mach. Intell. 23(6), 681–685 (2001).
[CrossRef]

Fabritius, T.

Fieguth, P. W.

A. Mishra, A. Wong, D. A. Clausi, and P. W. Fieguth, “Quasi-random nonlinear scale space,” Pattern Recognit. Lett. In Press. (Corrected Proof).

Garvin, M. M. K.

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

Guy, C. L.

M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
[CrossRef] [PubMed]

Jones, S. M.

Joo, C.

Kaplan, F.

M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
[CrossRef] [PubMed]

Kardon, R. R.

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

Kingsbury, N. G.

I. W. Selesnick, R. G. Baraniuk, and N. G. Kingsbury, “The Dual-Tree Complex Wavelet Transform,” IEEE Signal Process. Mag. 22(6), 123–151 (2005).
[CrossRef]

Koozekanani, D.

D. Koozekanani, K. Boyer, and C. Roberts, “Retinal thickness measurements from optical coherence tomography using a Markov boundary model,” IEEE Trans. Med. Imaging 20(9), 900–916 (2001).
[CrossRef] [PubMed]

Kopka, J.

M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
[CrossRef] [PubMed]

Makita, S.

Mishra, A.

Miura, M.

Mujat, M.

Myllylä, R.

Oliver, S. S.

Park, B.

Puliafito, C. A.

Roberts, C.

D. Koozekanani, K. Boyer, and C. Roberts, “Retinal thickness measurements from optical coherence tomography using a Markov boundary model,” IEEE Trans. Med. Imaging 20(9), 900–916 (2001).
[CrossRef] [PubMed]

Russell, S. S. R.

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

Salinas, H. M.

Scholz, M.

M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
[CrossRef] [PubMed]

Selbig, J.

M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
[CrossRef] [PubMed]

Selesnick, I. W.

I. W. Selesnick, R. G. Baraniuk, and N. G. Kingsbury, “The Dual-Tree Complex Wavelet Transform,” IEEE Signal Process. Mag. 22(6), 123–151 (2005).
[CrossRef]

Sonka, M. M.

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

Sorzano, C. O. S.

C. O. S. Sorzano, P. Thevenaz, and M. Unser, “Elastic registration of biological images using vector-spline regularization,” BIEEE Biomed. Eng. 52(4), 652–663 (2005).
[CrossRef]

Taylor, C. J.

T. F. Cootes, G. J. Edwards, and C. J. Taylor, “Active Appearance Models,” IEEE Trans. Pattern Anal. Mach. Intell. 23(6), 681–685 (2001).
[CrossRef]

Thevenaz, P.

C. O. S. Sorzano, P. Thevenaz, and M. Unser, “Elastic registration of biological images using vector-spline regularization,” BIEEE Biomed. Eng. 52(4), 652–663 (2005).
[CrossRef]

Unser, M.

C. O. S. Sorzano, P. Thevenaz, and M. Unser, “Elastic registration of biological images using vector-spline regularization,” BIEEE Biomed. Eng. 52(4), 652–663 (2005).
[CrossRef]

Werner, J. S.

Wong, A.

Wu, X. X.

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

Yasuno, Y.

Zawadzki, R. J.

BIEEE Biomed. Eng.

C. O. S. Sorzano, P. Thevenaz, and M. Unser, “Elastic registration of biological images using vector-spline regularization,” BIEEE Biomed. Eng. 52(4), 652–663 (2005).
[CrossRef]

Bioinformatics

M. Scholz, F. Kaplan, C. L. Guy, J. Kopka, and J. Selbig, “Non-linear PCA: a missing data approach,” Bioinformatics 21(20), 3887–3895 (2005).
[CrossRef] [PubMed]

IEEE Signal Process. Mag.

I. W. Selesnick, R. G. Baraniuk, and N. G. Kingsbury, “The Dual-Tree Complex Wavelet Transform,” IEEE Signal Process. Mag. 22(6), 123–151 (2005).
[CrossRef]

IEEE Trans. Med. Imaging

M. M. K. Garvin, M. M. D. Abramoff, R. R. Kardon, S. S. R. Russell, X. X. Wu, and M. M. Sonka, “Intraretinal Layer Segmentation of Macular Optical Coherence Tomography Images Using Optimal 3-D Graph Search,” IEEE Trans. Med. Imaging 27(10), 1495–1505 (2008).
[CrossRef] [PubMed]

D. Koozekanani, K. Boyer, and C. Roberts, “Retinal thickness measurements from optical coherence tomography using a Markov boundary model,” IEEE Trans. Med. Imaging 20(9), 900–916 (2001).
[CrossRef] [PubMed]

IEEE Trans. Pattern Anal. Mach. Intell.

T. F. Cootes, G. J. Edwards, and C. J. Taylor, “Active Appearance Models,” IEEE Trans. Pattern Anal. Mach. Intell. 23(6), 681–685 (2001).
[CrossRef]

J. Opt. Soc. Am. A

Opt. Express

Pattern Recognit. Lett.

A. Mishra, A. Wong, D. A. Clausi, and P. W. Fieguth, “Quasi-random nonlinear scale space,” Pattern Recognit. Lett. In Press. (Corrected Proof).

Other

A. A. Efros, and W. T. Freeman, “Image quilting for texture synthesis and transfer,” in Proceedings of the 28th annual conference on Computer graphics and interactive techniques(ACM, 2001), pp. 341–346.

M. Scholz, M. Fraunholz, and J. Selbig, “Nonlinear Principal Component Analysis: Neural Network Models and Applications,” in Principal Manifolds for Data Visualization and Dimension Reduction(2007), pp. 44–67.

A. K. Mishra, P. W. Fieguth, and D. A. Clausi, “Decoupled Active Contour (DAC) for Boundary Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence 99.

D. Tolliver, Y. Koutis, H. Ishikawa, J. S. Schuman, and G. L. Miller, “Unassisted Segmentation of Multiple Retinal Layers via Spectral Rounding,” in ARVO(2008).

P. Thevenaz, and M. Unser, “A pyramid approach to sub-pixel image fusion based on mutual information,” in Image Processing, 1996. Proceedings., International Conference on(1996), p. 265.

W. Drexler, and J. G. Fujimoto, Optical Coherence Tomography: Technology and Applications (Springer, 2008).

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 (9)

Fig. 1
Fig. 1

Algorithm overview: manually segmented data is used as the input to the training phase of the algorithm. After passing the pre-processing block a statistical model is constructed that captures the variance in the training data, which can be then used to segment unseen data.

Fig. 2
Fig. 2

Initial segmentation step of a despeckled OCT frame (on the left) after adaptive thresholding boundary detection demarking (on the right): internal limiting membrane (ILM, red), connecting cilia (CL, blue), retinal pigment epithelium (RPE, green).

Fig. 3
Fig. 3

Filling the gaps after the registration with NaNs and applying inverse neural network nonlinear PCA dimensionality reduction. In the case of the example data shown on the right, we can see that already the first eigenvector (e1) captures most of the variance in the original data set. This illustrates the idea behind the dimensionality reduction.

Fig. 4
Fig. 4

On the left is the result after the global low-res optimisation followed by, on the right, the refined result by the A-scan optimization.

Fig. 5
Fig. 5

Robust performance for all the layers is achieved even in presence of shadowing. A despeckled image is shown on the left; the segmented image is on the right.

Fig. 6
Fig. 6

Median and coefficient of variation computed on thickness maps of all the individual layers (nerve fibre layer (NFL), ganglion cell layer and inner plexiform layer (GCL + IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), connecting cilia (CL), outer segment (OS), retinal pigment epithelium (RPE)), as well as the retina, obtained from 17 eyes.

Fig. 7
Fig. 7

Segmentation in a case of added strong noise. Left original image. Right filtered, denoised image with segmentation results superimposed.

Fig. 8
Fig. 8

Error rates E B (Basic) and E L D E V (Layer DEViation) with decreasing dynamic range for all the data sets, with confidence interval (1.96 * standard deviation) marked by the dashed lines. For both error measures a slow rise in the error values can be observed, which guarantees robust performance with noisy data.

Fig. 9
Fig. 9

Error rates for the individual layers E B (Basic) and E L D E V (Layer DEViation) with decreasing dynamic range for all the data sets. For all the individual layers (nerve fibre layer (NFL), ganglion cell layer and inner plexiform layer (GCL + IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), connecting cilia (CL), outer segment (OS), retinal pigment epithelium (RPE)) a slow rise in the error values can be observed. Thin layers inherently exhibit greater error values, as both errors are normalized by the layer area.

Tables (3)

Tables Icon

Table 1 Variability of manual segmentations on 75 B-scans in percent (the data has been previously examined and “bad” results left out)

Tables Icon

Table 2 Error values on 466 B-scans at various positions from 17 eyes in percent before the A-scan optimization

Tables Icon

Table 3 Error values on 466 B-scans at various positions from 17 eyes in percent after the A-scan optimization

Equations (12)

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

X = ( x 1 x m ) = ( v 11 v 1 n v m 1 v m n ) v i j = [ off 1 off W ]
X = W Σ V T Y = W L T X
Φ g e n : z > X ^ Φ e x t r : X > z
x ( x  -  μ ( x ) 1 )/ σ ( x )
g = G ( x , I )
x = S ( s ) = x ¯ + Q s s g = T ( t ) = g ¯ + Q g t
f ( s , t ) = | T 1 ( G ( S ( s ) , I ) ) t | + b * ( S ( s ) b b i n i t ) 2 w
A = ( a Off 11 a Off 1 n a Off u 1 a Off u n )
E i B = j = 1 j = w | y A u t i j y R e f i j | , E i L D E V = w * j = 1 j = w ( y A u t i j y R e f i j ) 2
E i B = j = 1 j = w | y A u t i j y R e f i j | = j = 1 j = w | d | = w * | d | E i L D E V = w * j = 1 j = w ( y A u t i j y R e f i j ) 2 = w * j = 1 j = w d 2 = w * w * d 2 = w * | d | = E i B
E = i = 1 i = b E i A
E = i = k i = k + 1 E i A A + A R , 0 < k < b

Metrics