A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure

Sandeep N. Menon; V. B. Vineeth Reddy; A. Yeshwanth; B. N. Anoop; Jeny Rajan

doi:10.1007/978-981-32-9291-8_10

A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure

Sandeep N. Menon, V. B. Vineeth Reddy, A. Yeshwanth, B. N. Anoop^*, Jeny Rajan

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

13 Citations (Scopus)

Abstract

Optical coherence tomography (OCT) is an imaging technique widely used to image retina. Speckle noise in OCT images generally degrades the quality of the OCT images and makes the clinical diagnosis tedious. This paper proposes a new deep neural network despeckling scheme called gated convolution–deconvolution structure (GCDS). The robustness of the proposed method is evaluated on the publicly available OPTIMA challenge dataset and Duke dataset. The quantitative analysis based on PSNR shows that the results of the proposed method are superior to other state-of-the-art methods. The application of the proposed method for segmenting retinal cyst from OPTIMA challenge dataset was also studied.

Original language	English
Title of host publication	Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018
Editors	Bidyut B. Chaudhuri, Masaki Nakagawa, Pritee Khanna, Sanjeev Kumar
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	115-126
Number of pages	12
ISBN (Print)	9789813292901
DOIs	https://doi.org/10.1007/978-981-32-9291-8_10
Publication status	Published - 2020
Event	3rd International Conference on Computer Vision and Image Processing, CVIP 2018 - Jabalpur, India Duration: 29-09-2019 → 01-10-2019

Publication series

Name	Advances in Intelligent Systems and Computing
Volume	1024
ISSN (Print)	2194-5357
ISSN (Electronic)	2194-5365

Conference

Conference	3rd International Conference on Computer Vision and Image Processing, CVIP 2018
Country/Territory	India
City	Jabalpur
Period	29-09-19 → 01-10-19

All Science Journal Classification (ASJC) codes

Control and Systems Engineering
General Computer Science

Access to Document

10.1007/978-981-32-9291-8_10

Cite this

Menon, S. N., Vineeth Reddy, V. B., Yeshwanth, A., Anoop, B. N., & Rajan, J. (2020). A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure. In B. B. Chaudhuri, M. Nakagawa, P. Khanna, & S. Kumar (Eds.), Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018 (pp. 115-126). (Advances in Intelligent Systems and Computing; Vol. 1024). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-32-9291-8_10

Menon, Sandeep N. ; Vineeth Reddy, V. B. ; Yeshwanth, A. et al. / A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure. Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018. editor / Bidyut B. Chaudhuri ; Masaki Nakagawa ; Pritee Khanna ; Sanjeev Kumar. Springer Science and Business Media Deutschland GmbH, 2020. pp. 115-126 (Advances in Intelligent Systems and Computing).

@inproceedings{c3180e60fe284eeb96ab01e0bbfa2155,

title = "A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure",

abstract = "Optical coherence tomography (OCT) is an imaging technique widely used to image retina. Speckle noise in OCT images generally degrades the quality of the OCT images and makes the clinical diagnosis tedious. This paper proposes a new deep neural network despeckling scheme called gated convolution–deconvolution structure (GCDS). The robustness of the proposed method is evaluated on the publicly available OPTIMA challenge dataset and Duke dataset. The quantitative analysis based on PSNR shows that the results of the proposed method are superior to other state-of-the-art methods. The application of the proposed method for segmenting retinal cyst from OPTIMA challenge dataset was also studied.",

author = "Menon, \{Sandeep N.\} and \{Vineeth Reddy\}, \{V. B.\} and A. Yeshwanth and Anoop, \{B. N.\} and Jeny Rajan",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Singapore Pte Ltd.; 3rd International Conference on Computer Vision and Image Processing, CVIP 2018 ; Conference date: 29-09-2019 Through 01-10-2019",

year = "2020",

doi = "10.1007/978-981-32-9291-8\_10",

language = "English",

isbn = "9789813292901",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "115--126",

editor = "Chaudhuri, \{Bidyut B.\} and Masaki Nakagawa and Pritee Khanna and Sanjeev Kumar",

booktitle = "Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018",

address = "Germany",

}

Menon, SN, Vineeth Reddy, VB, Yeshwanth, A, Anoop, BN & Rajan, J 2020, A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure. in BB Chaudhuri, M Nakagawa, P Khanna & S Kumar (eds), Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018. Advances in Intelligent Systems and Computing, vol. 1024, Springer Science and Business Media Deutschland GmbH, pp. 115-126, 3rd International Conference on Computer Vision and Image Processing, CVIP 2018, Jabalpur, India, 29-09-19. https://doi.org/10.1007/978-981-32-9291-8_10

A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure. / Menon, Sandeep N.; Vineeth Reddy, V. B.; Yeshwanth, A. et al.
Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018. ed. / Bidyut B. Chaudhuri; Masaki Nakagawa; Pritee Khanna; Sanjeev Kumar. Springer Science and Business Media Deutschland GmbH, 2020. p. 115-126 (Advances in Intelligent Systems and Computing; Vol. 1024).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure

AU - Menon, Sandeep N.

AU - Vineeth Reddy, V. B.

AU - Yeshwanth, A.

AU - Anoop, B. N.

AU - Rajan, Jeny

PY - 2020

Y1 - 2020

N2 - Optical coherence tomography (OCT) is an imaging technique widely used to image retina. Speckle noise in OCT images generally degrades the quality of the OCT images and makes the clinical diagnosis tedious. This paper proposes a new deep neural network despeckling scheme called gated convolution–deconvolution structure (GCDS). The robustness of the proposed method is evaluated on the publicly available OPTIMA challenge dataset and Duke dataset. The quantitative analysis based on PSNR shows that the results of the proposed method are superior to other state-of-the-art methods. The application of the proposed method for segmenting retinal cyst from OPTIMA challenge dataset was also studied.

AB - Optical coherence tomography (OCT) is an imaging technique widely used to image retina. Speckle noise in OCT images generally degrades the quality of the OCT images and makes the clinical diagnosis tedious. This paper proposes a new deep neural network despeckling scheme called gated convolution–deconvolution structure (GCDS). The robustness of the proposed method is evaluated on the publicly available OPTIMA challenge dataset and Duke dataset. The quantitative analysis based on PSNR shows that the results of the proposed method are superior to other state-of-the-art methods. The application of the proposed method for segmenting retinal cyst from OPTIMA challenge dataset was also studied.

UR - https://www.scopus.com/pages/publications/85075609908

UR - https://www.scopus.com/inward/citedby.url?scp=85075609908&partnerID=8YFLogxK

U2 - 10.1007/978-981-32-9291-8_10

DO - 10.1007/978-981-32-9291-8_10

M3 - Conference contribution

AN - SCOPUS:85075609908

SN - 9789813292901

T3 - Advances in Intelligent Systems and Computing

SP - 115

EP - 126

BT - Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018

A2 - Chaudhuri, Bidyut B.

A2 - Nakagawa, Masaki

A2 - Khanna, Pritee

A2 - Kumar, Sanjeev

PB - Springer Science and Business Media Deutschland GmbH

T2 - 3rd International Conference on Computer Vision and Image Processing, CVIP 2018

Y2 - 29 September 2019 through 1 October 2019

ER -

Menon SN, Vineeth Reddy VB, Yeshwanth A, Anoop BN, Rajan J. A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure. In Chaudhuri BB, Nakagawa M, Khanna P, Kumar S, editors, Proceedings of 3rd International Conference on Computer Vision and Image Processing, CVIP 2018. Springer Science and Business Media Deutschland GmbH. 2020. p. 115-126. (Advances in Intelligent Systems and Computing). doi: 10.1007/978-981-32-9291-8_10

A Novel Deep Learning Approach for the Removal of Speckle Noise from Optical Coherence Tomography Images Using Gated Convolution–Deconvolution Structure

Abstract

Publication series

Conference

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this