EffUnet-SpaGen: An Efficient and Spatial Generative Approach to Glaucoma Detection

Venkatesh Krishna Adithya; Bryan M. Williams; Silvester Czanner; Srinivasan Kavitha; David S. Friedman; Colin E. Willoughby; Rengaraj Venkatesh; Gabriela Czanner

doi:10.3390/jimaging7060092

EffUnet-SpaGen: An Efficient and Spatial Generative Approach to Glaucoma Detection

Venkatesh Krishna Adithya
, Bryan M. Williams
, Silvester Czanner
, Srinivasan Kavitha
, David S. Friedman
, Colin E. Willoughby
, Rengaraj Venkatesh
, Gabriela Czanner

Centre for Genomic Medicine

Research output: Contribution to journal › Article › peer-review

24 Citations (Scopus)

97 Downloads (Pure)

Abstract

Current research in automated disease detection focuses on making algorithms “slimmer” reducing the need for large training datasets and accelerating recalibration for new data while achieving high accuracy. The development of slimmer models has become a hot research topic in medical imaging. In this work, we develop a two-phase model for glaucoma detection, identifying and exploiting a redundancy in fundus image data relating particularly to the geometry. We propose a novel algorithm for the cup and disc segmentation “EffUnet” with an efficient convolution block and combine this with an extended spatial generative approach for geometry modelling and classification, termed “SpaGen” We demonstrate the high accuracy achievable by EffUnet in detecting the optic disc and cup boundaries and show how our algorithm can be quickly trained with new data by recalibrating the EffUnet layer only. Our resulting glaucoma detection algorithm, “EffUnet-SpaGen”, is optimized to significantly reduce the computational burden while at the same time surpassing the current state-of-art in glaucoma detection algorithms with AUROC 0.997 and 0.969 in the benchmark online datasets ORIGA and DRISHTI, respectively. Our algorithm also allows deformed areas of the optic rim to be displayed and investigated, providing explainability, which is crucial to successful adoption and implementation in clinical settings.

Original language	English
Article number	92
Number of pages	17
Journal	Journal of Imaging
Volume	7
Issue number	6
Early online date	30 May 2021
DOIs	https://doi.org/10.3390/jimaging7060092
Publication status	Published (in print/issue) - 30 Jun 2021

Bibliographical note

Funding Information:
Acknowledgments: V.A.K., G.C. and S.C. are thankful for their support from Global Challenges Research Fund to Liverpool John Moores University.

Funding Information:
The ORIGA dataset is a subset of the data from the Singapore Malay Eye Study (SiMES), collected from 2004 to 2007 by the Singapore Eye Research Institute and funded by the National Medical Research Council. All images were anonymized before release. The ORIGA dataset comprises 482 healthy and 168 glaucoma images from Malay adults aged 40–80. The 650 images with manually labelled optic masks are divided into 325 training images (including 72 glaucoma cases), called ORIGA-A, and 325 testing images (including 95 glaucoma cases), called ORIGA-B [29]. The images were manually annotated by an ophthalmologist clicking on several locations of the image to indicate the optic disc and optic rim, then a best-fitting ellipse was calculated automatically. We refer to this segmentation as the ground truth. Four graders also graded the image, and a fifth grader was used for consensus.

Publisher Copyright:
© 2021 by the authors. Licensee MDPI, Basel, Switzerland.

Funding

Funding Information: Acknowledgments: V.A.K., G.C. and S.C. are thankful for their support from Global Challenges Research Fund to Liverpool John Moores University. Funding Information: The ORIGA dataset is a subset of the data from the Singapore Malay Eye Study (SiMES), collected from 2004 to 2007 by the Singapore Eye Research Institute and funded by the National Medical Research Council. All images were anonymized before release. The ORIGA dataset comprises 482 healthy and 168 glaucoma images from Malay adults aged 40–80. The 650 images with manually labelled optic masks are divided into 325 training images (including 72 glaucoma cases), called ORIGA-A, and 325 testing images (including 95 glaucoma cases), called ORIGA-B [29]. The images were manually annotated by an ophthalmologist clicking on several locations of the image to indicate the optic disc and optic rim, then a best-fitting ellipse was calculated automatically. We refer to this segmentation as the ground truth. Four graders also graded the image, and a fifth grader was used for consensus. Publisher Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Classification
Diagnosis
Generative model
Glaucoma
Machine learning

Access to Document

10.3390/jimaging7060092Licence: CC BY

jimaging-07-00092-v2Final published version, 1.74 MBLicence: CC BY

Cite this

@article{d10c5128d39c43ea9e8f5db3dd2ebf43,

title = "EffUnet-SpaGen: An Efficient and Spatial Generative Approach to Glaucoma Detection",

abstract = "Current research in automated disease detection focuses on making algorithms “slimmer” reducing the need for large training datasets and accelerating recalibration for new data while achieving high accuracy. The development of slimmer models has become a hot research topic in medical imaging. In this work, we develop a two-phase model for glaucoma detection, identifying and exploiting a redundancy in fundus image data relating particularly to the geometry. We propose a novel algorithm for the cup and disc segmentation “EffUnet” with an efficient convolution block and combine this with an extended spatial generative approach for geometry modelling and classification, termed “SpaGen” We demonstrate the high accuracy achievable by EffUnet in detecting the optic disc and cup boundaries and show how our algorithm can be quickly trained with new data by recalibrating the EffUnet layer only. Our resulting glaucoma detection algorithm, “EffUnet-SpaGen”, is optimized to significantly reduce the computational burden while at the same time surpassing the current state-of-art in glaucoma detection algorithms with AUROC 0.997 and 0.969 in the benchmark online datasets ORIGA and DRISHTI, respectively. Our algorithm also allows deformed areas of the optic rim to be displayed and investigated, providing explainability, which is crucial to successful adoption and implementation in clinical settings.",

keywords = "Classification, Diagnosis, Generative model, Glaucoma, Machine learning",

author = "\{Krishna Adithya\}, Venkatesh and Williams, \{Bryan M.\} and Silvester Czanner and Srinivasan Kavitha and Friedman, \{David S.\} and Willoughby, \{Colin E.\} and Rengaraj Venkatesh and Gabriela Czanner",

note = "Funding Information: Acknowledgments: V.A.K., G.C. and S.C. are thankful for their support from Global Challenges Research Fund to Liverpool John Moores University. Funding Information: The ORIGA dataset is a subset of the data from the Singapore Malay Eye Study (SiMES), collected from 2004 to 2007 by the Singapore Eye Research Institute and funded by the National Medical Research Council. All images were anonymized before release. The ORIGA dataset comprises 482 healthy and 168 glaucoma images from Malay adults aged 40–80. The 650 images with manually labelled optic masks are divided into 325 training images (including 72 glaucoma cases), called ORIGA-A, and 325 testing images (including 95 glaucoma cases), called ORIGA-B [29]. The images were manually annotated by an ophthalmologist clicking on several locations of the image to indicate the optic disc and optic rim, then a best-fitting ellipse was calculated automatically. We refer to this segmentation as the ground truth. Four graders also graded the image, and a fifth grader was used for consensus. Publisher Copyright: {\textcopyright} 2021 by the authors. Licensee MDPI, Basel, Switzerland.",

year = "2021",

month = jun,

day = "30",

doi = "10.3390/jimaging7060092",

language = "English",

volume = "7",

journal = "Journal of Imaging",

issn = "2313-433X",

publisher = "Multidisciplinary Digital Publishing Institute",

number = "6",

}

TY - JOUR

T1 - EffUnet-SpaGen: An Efficient and Spatial Generative Approach to Glaucoma Detection

AU - Krishna Adithya, Venkatesh

AU - Williams, Bryan M.

AU - Czanner, Silvester

AU - Kavitha, Srinivasan

AU - Friedman, David S.

AU - Willoughby, Colin E.

AU - Venkatesh, Rengaraj

AU - Czanner, Gabriela

N1 - Funding Information: Acknowledgments: V.A.K., G.C. and S.C. are thankful for their support from Global Challenges Research Fund to Liverpool John Moores University. Funding Information: The ORIGA dataset is a subset of the data from the Singapore Malay Eye Study (SiMES), collected from 2004 to 2007 by the Singapore Eye Research Institute and funded by the National Medical Research Council. All images were anonymized before release. The ORIGA dataset comprises 482 healthy and 168 glaucoma images from Malay adults aged 40–80. The 650 images with manually labelled optic masks are divided into 325 training images (including 72 glaucoma cases), called ORIGA-A, and 325 testing images (including 95 glaucoma cases), called ORIGA-B [29]. The images were manually annotated by an ophthalmologist clicking on several locations of the image to indicate the optic disc and optic rim, then a best-fitting ellipse was calculated automatically. We refer to this segmentation as the ground truth. Four graders also graded the image, and a fifth grader was used for consensus. Publisher Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland.

PY - 2021/6/30

Y1 - 2021/6/30

N2 - Current research in automated disease detection focuses on making algorithms “slimmer” reducing the need for large training datasets and accelerating recalibration for new data while achieving high accuracy. The development of slimmer models has become a hot research topic in medical imaging. In this work, we develop a two-phase model for glaucoma detection, identifying and exploiting a redundancy in fundus image data relating particularly to the geometry. We propose a novel algorithm for the cup and disc segmentation “EffUnet” with an efficient convolution block and combine this with an extended spatial generative approach for geometry modelling and classification, termed “SpaGen” We demonstrate the high accuracy achievable by EffUnet in detecting the optic disc and cup boundaries and show how our algorithm can be quickly trained with new data by recalibrating the EffUnet layer only. Our resulting glaucoma detection algorithm, “EffUnet-SpaGen”, is optimized to significantly reduce the computational burden while at the same time surpassing the current state-of-art in glaucoma detection algorithms with AUROC 0.997 and 0.969 in the benchmark online datasets ORIGA and DRISHTI, respectively. Our algorithm also allows deformed areas of the optic rim to be displayed and investigated, providing explainability, which is crucial to successful adoption and implementation in clinical settings.

AB - Current research in automated disease detection focuses on making algorithms “slimmer” reducing the need for large training datasets and accelerating recalibration for new data while achieving high accuracy. The development of slimmer models has become a hot research topic in medical imaging. In this work, we develop a two-phase model for glaucoma detection, identifying and exploiting a redundancy in fundus image data relating particularly to the geometry. We propose a novel algorithm for the cup and disc segmentation “EffUnet” with an efficient convolution block and combine this with an extended spatial generative approach for geometry modelling and classification, termed “SpaGen” We demonstrate the high accuracy achievable by EffUnet in detecting the optic disc and cup boundaries and show how our algorithm can be quickly trained with new data by recalibrating the EffUnet layer only. Our resulting glaucoma detection algorithm, “EffUnet-SpaGen”, is optimized to significantly reduce the computational burden while at the same time surpassing the current state-of-art in glaucoma detection algorithms with AUROC 0.997 and 0.969 in the benchmark online datasets ORIGA and DRISHTI, respectively. Our algorithm also allows deformed areas of the optic rim to be displayed and investigated, providing explainability, which is crucial to successful adoption and implementation in clinical settings.

KW - Classification

KW - Diagnosis

KW - Generative model

KW - Glaucoma

KW - Machine learning

UR - https://www.scopus.com/pages/publications/85108272370

U2 - 10.3390/jimaging7060092

DO - 10.3390/jimaging7060092

M3 - Article

SN - 2313-433X

VL - 7

JO - Journal of Imaging

JF - Journal of Imaging

IS - 6

M1 - 92

ER -

EffUnet-SpaGen: An Efficient and Spatial Generative Approach to Glaucoma Detection

Abstract

Bibliographical note

Funding

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this