Your browser does not support JavaScript!
http://iet.metastore.ingenta.com
1887

access icon free Deblurring retinal optical coherence tomography via a convolutional neural network with anisotropic and double convolution layer

  • Author(s): Jian Lian 1, 2, 3, 4 ; Sujuan Hou 2, 3, 4 ; Xiaodan Sui 2, 3, 4 ; Fangzhou Xu 5 ; Yuanjie Zheng 2, 3, 4
    • View affiliations
  • Source: Volume 12, Issue 6, September 2018, p. 900 – 907
    DOI:  10.1049/iet-cvi.2018.0016 , Print ISSN 1751-9632, Online ISSN 1751-9640
© The Institution of Engineering and Technology
Received 11/01/2018, Accepted 18/04/2018, Revised 22/03/2018, Published 26/04/2018

Various image pre-processing tasks in optical coherence tomography (OCT) systems involve reversing degradation effects (e.g. deblurring). Current deblurring research mainly focuses on how to build suitable degradation models using deconvolution operators. However, model-based solutions may not work well in many scenarios. To solve this problem, the authors propose a non-model architecture, called a deep convolutional neural network, to address parameter-free situations. The proposed solution employs a deep learning strategy to bridge the gap between traditional model-based methods and neural network architectures. Experiments on retinal OCT images demonstrate that the proposed approach achieves superior performance compared with the state-of-the-art model-based OCT deblurring methods.

Inspec keywords: neural nets; biomedical optical imaging; convolution; learning (artificial intelligence); image restoration; deconvolution; medical image processing; eye; optical tomography

Other keywords: traditional model; current deblurring research; OCT deblurring methods; degradation effects; suitable degradation models; deep learning strategy; retinal OCT images; neural network architectures; deconvolution operators; deep convolutional neural network; parameter-free situations; image pre-processing tasks; double convolution layer; state-of-the-art model; model-based solutions; retinal optical coherence tomography; nonmodel architecture; optical coherence tomography systems

Subjects: Computer vision and image processing techniques; Optical, image and video signal processing; Patient diagnostic methods and instrumentation; Optical and laser radiation (medical uses)

References

    1. 1)
      • 32. LeCun, Y., Bottou, L., Bengio, Y., et al: ‘Gradient-based learning applied to document recognition’. Proc. IEEE, 1998, 86, (11), pp. 2278–2324.
    2. 2)
      • 53. Boscaini, D.: ‘Learning shape correspondence with anisotropic convolutional neural networks’. arXiv preprint, 2016, arXiv 1605.06437.
    3. 3)
      • 5. Nagy, Z.Z., Ecsedy, M., Kovcs, I., et al: ‘Macular morphology assessed by optical coherence tomography image segmentation after femtosecond laser-assisted and standard cataract surgery’, J. Cataract Refract. Surg., 2012, 38, (6), pp. 941–946.
    4. 4)
      • 2. Schmitt, J.M.: ‘Optical coherence tomography (OCT): a review’, IEEE J. Sel. Top. Quantum Electron., 2002, 5, (4), pp. 1205–1215.
    5. 5)
      • 41. Weston, J., Bengio, S., Usunier, N.: ‘Large scale image annotation: learning to rank with joint word-image embeddings’, Mach. Learn., 2010, 81, (1), pp. 21–35.
    6. 6)
      • 26. Kenig, T., Kam, Z., Feuer, A.: ‘Blind image deconvolution using machine learning for three-dimensional microscopy’, IEEE Trans. Pattern Anal. Mach. Intell., 2010, 32, (12), pp. 2191–2204.
    7. 7)
      • 21. Cho, S., Wang, J., Lee, S.: ‘Handling outliers in non-blind image deconvolution’. IEEE Proc. of the Int. Conf. on Computer Vision, Barcelona, Spain, 2011, pp. 495–502.
    8. 8)
      • 27. Ralston, T.S., Marks, D.L., Kamalabadi, F., et al: ‘Gaussian beam deconvolution in optical coherence tomography’. Proc. of SPIE, San Diego, CA, USA, 2005, pp. 1–12.
    9. 9)
      • 8. Labiau, S., David, G., Gigan, S., et al: ‘Defocus test and defocus correction in full-field optical coherence tomography’, Opt. Lett., 2009, 34, (10), pp. 1576–1578.
    10. 10)
      • 43. Jia, Y., Shelhamer, E., Donahue, J., et al: ‘Caffe: convolutional architecture for fast feature embedding’. Proc. of the 22nd ACM Int. Conf. on Multimedia, Orlando, FL, USA, 2014, pp. 675–678.
    11. 11)
      • 15. Levin, A., Fergus, R., Durand, F., et al: ‘Deconvolution using natural image priors’, ACM Trans. Graph., 2007, 26, pp. 0–2.
    12. 12)
      • 13. Hammer, D.X., Ferguson, R.D., Magill, J.C., et al: ‘Active retinal tracker for clinical optical coherence tomography systems’, J. Biomed. Opt., 2005, 10, (2), pp. 02403800–02403811.
    13. 13)
      • 11. Potsaid, B., Gorczynska, I., Srinivasan, V.J., et al: ‘Ultrahigh speed spectral/Fourier domain oct ophthalmic imaging at 70,000 to 312,500 axial scans per second’, Opt. Express, 2008, 16, (19), pp. 15149–15169.
    14. 14)
      • 54. Savakis, A.E., Trussell, H.J.: ‘On the accuracy of PSF representation in image restoration’, IEEE Trans. Image Process., 1993, 2, (2), pp. 252–259.
    15. 15)
      • 44. Molchanov, P.: ‘Online detection and classification of dynamic hand gestures with recurrent 3D convolutional neural network’. Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 2016, pp. 4207–4215.
    16. 16)
      • 38. Chan, W., Jaitly, N., Le, Q.: ‘Listen, attend and spell: A neural network for large vocabulary conversational speech recognition’. Proc. of IEEE Int. Conf. on Acoustics, Speech and Signal Processing, Shanghai, China, 2016, pp. 4960–4964.
    17. 17)
      • 37. Deng, L., Seltzer, M.L., Yu, D., et al: ‘Binary coding of speech spectrograms using a deep auto-encoder’. Proc. of Interspeech, Makuhari, Japan, 2010, pp. 1692–1695.
    18. 18)
      • 36. Dahl, G.E., Yu, D., Deng, L., et al: ‘Context-dependent pre-trained deep neural networks for large-vocabulary speech recognition’, IEEE Trans. Audio, Speech, Lang. Process., 2012, 20, (1), pp. 30–42.
    19. 19)
      • 45. Krizhevsky, A., Sutskever, I., Hinton, G.E.: ‘Imagenet classification with deep convolutional neural networks’. Proc. of Advances in Neural Information Processing Systems, Lake Tahoe, NV, USA, 2012, pp. 1097–1105.
    20. 20)
      • 31. Wieschollek, P., Schölkopf, B., Lensch, H.P.A., et al: ‘End-to-end learning for image burst deblurring’. Asian Conf. on Computer Vision, 2016, pp. 35–51.
    21. 21)
      • 14. Zhang, Y., Yu, C., David, D.L.: ‘Optimizing Laguerre expansion based deconvolution methods for analysing bi-exponential fluorescence lifetime images’, Opt. Express, 2016, 24, (13), pp. 13894–13905.
    22. 22)
      • 42. Simonyan, K., Zisserman, A.: ‘Very deep convolutional networks for large-scale image recognition’, 2014, arXiv preprint arXiv:1409.1556.
    23. 23)
      • 23. Ahmed, A., Recht, B., Romberg, J.: ‘Blind deconvolution using convex programming’, IEEE Trans. Inf. Theory, 2014, 60, (3), pp. 1711–1732.
    24. 24)
      • 49. Turaga, S.C., Murray, J.F., Jain, V., et al: ‘Convolutional networks can learn to generate affinity graphs for image segmentation’, Neural Comput., 2010, 22, (2), pp. 511–538.
    25. 25)
      • 16. Ruiz, P., Zhou, X., Mateos, J., et al: ‘Variational Bayesian blind image deconvolution: a review’, Digit. Signal Process., 2015, 47, pp. 116–127.
    26. 26)
      • 24. Perrone, D.: ‘Total variation blind deconvolution: the devil is in the details’. Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, Boston, MA, USA, 2014, pp. 2909–2916.
    27. 27)
      • 22. Kundur, D., Hatzinakos, D.: ‘Blind image deconvolution’, IEEE Signal Process. Mag., 1996, 13, (3), pp. 43–64.
    28. 28)
      • 47. Simonyan, K., Vedaldi, A., Zisserman, A.: ‘Deep inside convolutional networks: visualising image classification models and saliency maps’, 2013, arXiv preprint arXiv:1312.6034.
    29. 29)
      • 10. Robinson, D.: ‘The mechanics of human saccadic eye movement’, J. Physiol., 1964, 174, (2), pp. 245–264.
    30. 30)
      • 3. Iester, M., Cordano, C., Costa, A.D., et al: ‘Effectiveness of time domain and spectral domain optical coherence tomograph to evaluate eyes with and, without optic neuritis in multiple sclerosi patients’, J Mult. Scler., 2016, 3, (173), pp. 1–4.
    31. 31)
      • 7. Liu, Y., Liang, Y., Mu, G., et al: ‘Deconvolution methods for image deblurring in optical coherence tomography’, J. Opt. Soc. Am. A, Opt. Image Sci. Vis., 2009, 26, (1), pp. 72–77.
    32. 32)
      • 17. Wiener, N.: ‘Smoothing of stationary time series’, Eng. Appl., (Wiley, New York, 1949).
    33. 33)
      • 20. Lee, J., Ho, Y.: ‘High-quality non-blind image deconvolution with adaptive regularization’, J. Vis. Commun. Image Represent., 2011, 22, (7), pp. 653–663.
    34. 34)
      • 52. Jain, V., Seung, S.: ‘Natural image denoising with convolutional networks’. Proc. of Advances in Neural Information Processing Systems (NIPS), Vancouver, Canada, 2009, pp. 769–776.
    35. 35)
      • 4. Kanagasingam, Y., Bhuiyan, A., Abramoff, M.D., et al: ‘Progress on retinal image analysis for age related macular degeneration’, Prog. Retin. Eye Res., 2014, 38, (1), pp. 20–42.
    36. 36)
      • 46. Socher, R., Huval, B., Bath, B., et al: ‘Convolutional-recursive deep learning for 3d object classification’. Proc. of Advances in Neural Information Processing Systems, Lake Tahoe, NV, USA, 2012, pp. 665–673.
    37. 37)
      • 1. Huang, D., Swanson, E.A., Lin, C.P., et al: ‘Optical coherence tomography’, Science, 1991, 254, (5035), pp. 1178–1181.
    38. 38)
      • 9. Kraus, M.F., Potsaid, B., Mayer, M.A., et al: ‘Motion correction in optical coherence tomography volumes on a per a-scan basis using orthogonal scan patterns’, Biomed. Opt. Express, 2012, 3, (6), pp. 1182–1199.
    39. 39)
      • 19. Fergus, R., William, T.: ‘Removing camera shake from a single photograph’. Proc. of ACM Transactions on Graphics, Boston, MA, USA, 2006, pp. 787–794.
    40. 40)
      • 29. Hojjatoleslami, S., Avanaki, M., Podoleanu, A.G.: ‘Image quality improvement in optical coherence tomography using Lucy-Richardson deconvolution algorithm’, Appl. Opt., 2013, 52, (23), pp. 5663–5670.
    41. 41)
      • 35. Dahl, G., Mohamed, A.R., Hinton, G.E.: ‘Phone recognition with the mean-covariance restricted Boltzmann machine’. Proc. of Advances in Neural Information Processing Systems, Vancouver, Canada, 2010, pp. 469–477.
    42. 42)
      • 40. Collobert, R., Weston, J., Bottou, L., et al: ‘Natural language processing (almost) from scratch’, J. Mach. Learn. Res., 2011, 12, (1), pp. 2493–2537.
    43. 43)
      • 28. Wang, K., Ding, Z., Chen, M., et al: ‘Deconvolution with fall-off compensated axial point spread function in spectral domain optical coherence tomography’, Opt. Commun., 2011, 284, (12), pp. 3173–3180.
    44. 44)
      • 58. Lian, J., Zheng, Y., Jiao, W., et al: ‘Deblurring sequential ocular images from multi-spectral imaging (MSI) via mutual information’, Med. Biol. Eng. Comput., 2017, 6, pp. 1–7.
    45. 45)
      • 48. Karpathy, A., Toderici, G., Shetty, S., et al: ‘Large-scale video classification with convolutional neural networks’. Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, Columbus, OH, USA, 2014, pp. 1725–1732.
    46. 46)
      • 12. Ferguson, R.D., Hammer, D.X., Paunescu, L.A., et al: ‘Tracking optical coherence tomography’, Opt. Lett., 2004, 29, (18), pp. 2139–2141.
    47. 47)
      • 55. Zhang, P., Liu, H., Ding, Y.: ‘Dynamic bee colony algorithm based on multi-species co-evolution’, Appl. Intell., 2014, 40, (3), pp. 427–440.
    48. 48)
      • 57. Wang, J., Gong, B., Liu, H., et al: ‘Multidisciplinary approaches to artificial swarm intelligence for heterogeneous computing and cloud scheduling’, Appl. Intell., 2015, 43, (3), pp. 662–675.
    49. 49)
      • 39. Boulanger-Lewandowski, N., Bengio, Y., Vincent, P.: ‘Modeling temporal dependencies in high-dimensional sequences: application to polyphonic music generation and transcription’, Chem. A, Eur. J., 2012, 18, (13), pp. 3981–3991.
    50. 50)
      • 18. Richardson, W.H.: ‘Bayesian-based iterative method of image restoration’, J. Opt. Soc. Am., 1972, 62, (1), pp. 55–59.
    51. 51)
      • 51. Long, J., Shelhamer, E., Darrell, T.: ‘Fully convolutional networks for semantic segmentation’. Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, Boston, MA, USA, 2015, pp. 3431–3440.
    52. 52)
      • 56. Steenken, S., Behrens, G., Schulte-Frohlinde, D.: ‘Heterogeneous computing and grid scheduling with parallel biologically inspired hybrid heuristics’, Trans. Inst. Meas. Control, 2014, 36, (6), pp. 805–814.
    53. 53)
      • 50. Prasoon, A., Petersen, K., Igel, C., et al: ‘Deep feature learning for knee cartilage segmentation using a triplanar convolutional neural network’. Proc. of Int. Conf. on Medical Image Computing and Computer-Assisted Intervention, 2013, pp. 246–253.
    54. 54)
      • 6. Spaide, R.F., Klancnik, J.M., Cooney, M.J.: ‘Retinal vascular layers imaged by fluorescein angiography and optical coherence tomography angiography’, JAMA Ophthalmol., 2015, 133, (1), pp. 45–50.
    55. 55)
      • 25. Schuler, C., Harold, C.B., Stefan, H.: ‘A machine learning approach for non-blind image deconvolution’. Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, Portland, OR, USA, 2013, pp. 1067–1074.
    56. 56)
      • 30. Xu, L., Ren, J.S.J., Liu, C., et al: ‘Deep convolutional neural network for image deconvolution’. Int. Conf. on Neural Information Processing Systems, Montreal, Canada, 2014, pp. 1790–1798.
    57. 57)
      • 34. Bengio, Y., Lamblin, P., Popovici, D., et al: ‘Greedy layer-wise training of deep networks’, Adv. Neural Inf. Process. Syst., 2007, 19, pp. 153–160.
    58. 58)
      • 33. Hinton, G.E., Osindero, S., Teh, Y.W.: ‘A fast learning algorithm for deep belief nets’, Neural Comput., 2006, 18, (7), pp. 1527–1554.
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-cvi.2018.0016
Loading

Related content

content/journals/10.1049/iet-cvi.2018.0016
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading
Print this page
This is a required field
Please enter a valid email address