Clinical Study | Published:

Supervised learning and dimension reduction techniques for quantification of retinal fluid in optical coherence tomography images

Eye volume 31, pages 12121220 (2017) | Download Citation

Abstract

Purpose

The purpose of the present study is to develop fast automated quantification of retinal fluid in optical coherence tomography (OCT) image sets.

Methods

We developed an image analysis pipeline tailored towards OCT images that consists of five steps for binary retinal fluid segmentation. The method is based on feature extraction, pre-segmention, dimension reduction procedures, and supervised learning tools.

Results

Fluid identification using our pipeline was tested on two separate patient groups: one associated to neovascular age-related macular degeneration, the other showing diabetic macular edema. For training and evaluation purposes, retinal fluid was annotated manually in each cross-section by human expert graders of the Vienna Reading Center. Compared with the manual annotations, our pipeline yields good quantification, visually and in numbers.

Conclusions

By demonstrating good automated retinal fluid quantification, our pipeline appears useful to expert graders within their current grading processes. Owing to dimension reduction, the actual learning part is fast and requires only few training samples. Hence, it is well-suited for integration into actual manufacturer’s devices, further improving segmentation by its use in daily clinical life.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

References

  1. 1.

    , , , . A view of the current and future role of optical coherence tomography in the management of age-related macular degeneration. Eye 2017; 31: 26–44.

  2. 2.

    , , . Deep learning. Nature 2015; 521: 436–444.

  3. 3.

    , , , , , et al. Automatic segmentation and classification of intraretinal cystoid fluid and subretinal fluid in 3D-OCT using convolutional neural networks. Invest Ophthalmol Vis Sci 2015; 560(7).

  4. 4.

    , , , , , . Computerized assessment of intraretinal and subretinal fluid regions in spectral-domain optical coherence tomography images of the retina. Am J Ophthalmol 2013; 155: 277–286.

  5. 5.

    , , , , , . 3D segmentation of fluid-associated abnormalities in retinal OCT: probability constrained graph-search—graph-cut. IEEE Trans Med Imaging 2012; 31(8): 1521–1531.

  6. 6.

    , , , , . Stratified sampling voxel classification for segmentation of intraretinal and subretinal fluid in longitudinal clinical OCT data. IEEE Trans Med Imaging 2015; 340(7): 1616–1623.

  7. 7.

    , , . Automated segmentation of intraretinal cystoid fluid in optical coherence tomography. IEEE Trans Biomed Eng 2012; 59(4): 1109–1114.

  8. 8.

    . Random walks for image segmentation. IEEE Trans Pattern Anal Mach Intell 2006; 28(11): 1768–1783.

  9. 9.

    , . Schrödinger Eigenmaps for the analysis of biomedical data. IEEE Trans Pattern Anal Mach Intell 2013; 35(5): 1274–1280.

  10. 10.

    , . Learning with Kernels. MIT Press: Cambridge, MA, 2002.

  11. 11.

    , . Nonlinear dimensionality reduction by locally linear embedding. Science 2000; 290(5500): 2323–2326.

  12. 12.

    . Complete discrete 2-D Gabor transforms by neural networks for image analysis and compression. IEEE Trans Acoust 1988; 360(7): 1169–1179.

  13. 13.

    , , . Nonlinear total variation based noise removal algorithm. Phys D 1992; 60: 259–268.

  14. 14.

    , . Fast partitioning of vector-valued images. SIAM J Imaging Sci 2014; 7: 1826–1852.

  15. 15.

    , , , , . Retinal vessel segmentation using the 2-D Gabor wavelet and supervised classification. IEEE Trans on Med Imaging 2006; 25(9): 1214–1222.

  16. 16.

    , . Optimal approximations by piecewise smooth functions and associated variational problems. Comm Pure Appl Math 1989; 420(5): 577–685.

  17. 17.

    . Some generalized order-disorder transformations. Proc Cambridge Philos Soc 1952; 48: 106–109.

  18. 18.

    . Pattern Recognition and Machine Learning. Springer-Verlag: New York, 2006.

  19. 19.

    , . Data Mining: Practical Machine Learning Tools and Techniques. Elsevier, 2005.

  20. 20.

    , , , , , . Automated 3-D intraretinal layer segmentation of macular spectral-domain optical coherence tomography images. IEEE Trans Med Imaging 2009; 28(9): 1436–1447.

  21. 21.

    , , , . Segmentation of the surfaces of the retinal layer from OCT images. Med Image Comput Comput Assist Interv 2006; 9(Pt 1): 800–807.

  22. 22.

    , , , , , et al. Correlation of 3-dimensionally quantified intraretinal and subretinal fluid with visual acuity in neovascular age-related macular degeneration. JAMA Ophthalmol 2016; 134(2): 182–190.

  23. 23.

    , , , , , et al. A novel benchmark model for intelligent annotation of spectral-domain optical coherence tomography scans using the example of cyst annotation. Comput Methods Programs Biomed 2016; 130: 93–105.

  24. 24.

    . Computerized interpretation of ECGs. Circ Arrhythm Electrophysiol 2013; 6: 2–4.

  25. 25.

    . Can we trust our ECG machines? Br J Cardiol 2014; 21: 62–63.

  26. 26.

    , , , . Shadow removal and contrast enhancement in optical coherence tomography images of the human optic nerve head. Invest Ophthalmol Vis Sci 2011; 52(10): 7738–7748.

  27. 27.

    , , , , . Automated vessel shadow segmentation of fovea-centered spectral-domain images from multiple OCT devices. Proc. SPIE 9034, Medical Imaging 2014; 9034: 903403.

  28. 28.

    , , , , , et al. Modeling photo-bleaching kinetics to create high resolution maps of rod rhodopsin in the human retina. PLoS One 2015; 10(7): e0131881.

  29. 29.

    , , . Deep learning is effective for classifying normal versus age-related macular degeneration optical coherence tomography images. Ophthalmol Retina, in press 2017.

Download references

Acknowledgements

This work was funded by the Vienna Science and Technology Fund (WWTF) through project VRG12-009.

Author information

Affiliations

  1. Department of Mathematics, University of Vienna, Vienna, Austria

    • A Breger
    •  & M Ehler
  2. Vienna Reading Center and Christian Doppler Laboratory for Ophthalmic Image Analysis, Department of Ophthalmology, Medical University of Vienna, Vienna, Austria

    • H Bogunovic
    • , S M Waldstein
    • , A-M Philip
    • , U Schmidt-Erfurth
    •  & B S Gerendas

Authors

  1. Search for A Breger in:

  2. Search for M Ehler in:

  3. Search for H Bogunovic in:

  4. Search for S M Waldstein in:

  5. Search for A-M Philip in:

  6. Search for U Schmidt-Erfurth in:

  7. Search for B S Gerendas in:

Competing interests

The authors declare no conflict of interest.

Corresponding author

Correspondence to B S Gerendas.

Supplementary information

About this article

Publication history

Received

Accepted

Published

DOI

https://doi.org/10.1038/eye.2017.61

Supplementary Information accompanies this paper on Eye website (http://www.nature.com/eye)