Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks

Muhammad Altaf Hussain; Syed Osama Bin Islam; M. I. Tiwana; Ubaid-Ur-Rehman; W. S. Qureshi

doi:10.1109/ICCAR.2019.8813469

Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks

Muhammad Altaf Hussain
, Syed Osama Bin Islam
, M. I. Tiwana
, Ubaid-Ur-Rehman
, W. S. Qureshi

National University of Sciences and Technology Pakistan

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

7 Citations (Scopus)

Abstract

Diabetic Retinopathy (DR) is an eye disorder that progressively leads to vision loss due to high glucose causing impairment of retinal blood vessels (BVs). 'Retinal Bright Lesions' such as 'Hard Exudates' (HEs) are plasma leakages from rapture retinal capillaries. HEs appear as hard, waxy, yellowish deposits from tiny spots to fat patches and signify moderate-severe Non-Proliferative Diabetic Retinopathy (NPDR). This paper proposes a simple, compact and computationally inexpensive technique for detection and classification of HEs using Digital Image Processing Techniques on digital fundus images complements and Artificial Neural Networks (ANN). The proposed technique unfolds through five stages i.e. Pre-processing, coarse detection, optimization, features detection extraction followed by classification. 'Speed Up Robust Features' (SURF) algorithm has been used for features detection extraction while 'Feed-Forward Back-propagation' (FFBP) ANN has been used for classification. The proposed technique has yielded 98.7% 'Sensitivity' (SE), 97.5% 'Specificity' (SP) and 97.7% 'Accuracy' (AC) on 'DIARETDB1' fundus images.

Original language	English
Title of host publication	2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019
Publisher	IEEE
Pages	206-211
Number of pages	6
ISBN (Electronic)	9781728133263
DOIs	https://doi.org/10.1109/ICCAR.2019.8813469
Publication status	Published - Apr 2019
Externally published	Yes
Event	5th International Conference on Control, Automation and Robotics, ICCAR 2019 - Beijing, China Duration: 19 Apr 2019 → 22 Apr 2019

Publication series

Name	2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019

Conference

Conference	5th International Conference on Control, Automation and Robotics, ICCAR 2019
Country/Territory	China
City	Beijing
Period	19/04/19 → 22/04/19

UN SDGs

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

SDG 3 Good Health and Well-being

Keywords

artificial neural networks
back-propagation
confusion matrix
diabetic retinopathy
green channel
hard exudates
image complements
optimization
thresholding

Access to Document

10.1109/ICCAR.2019.8813469

Cite this

Hussain, M. A., Islam, S. O. B., Tiwana, M. I., Ubaid-Ur-Rehman, & Qureshi, W. S. (2019). Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks. In 2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019 (pp. 206-211). Article 8813469 (2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019). IEEE. https://doi.org/10.1109/ICCAR.2019.8813469

@inproceedings{585da4af2dee41528f16988054b0af72,

title = "Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks",

abstract = "Diabetic Retinopathy (DR) is an eye disorder that progressively leads to vision loss due to high glucose causing impairment of retinal blood vessels (BVs). 'Retinal Bright Lesions' such as 'Hard Exudates' (HEs) are plasma leakages from rapture retinal capillaries. HEs appear as hard, waxy, yellowish deposits from tiny spots to fat patches and signify moderate-severe Non-Proliferative Diabetic Retinopathy (NPDR). This paper proposes a simple, compact and computationally inexpensive technique for detection and classification of HEs using Digital Image Processing Techniques on digital fundus images complements and Artificial Neural Networks (ANN). The proposed technique unfolds through five stages i.e. Pre-processing, coarse detection, optimization, features detection extraction followed by classification. 'Speed Up Robust Features' (SURF) algorithm has been used for features detection extraction while 'Feed-Forward Back-propagation' (FFBP) ANN has been used for classification. The proposed technique has yielded 98.7\% 'Sensitivity' (SE), 97.5\% 'Specificity' (SP) and 97.7\% 'Accuracy' (AC) on 'DIARETDB1' fundus images.",

keywords = "artificial neural networks, back-propagation, confusion matrix, diabetic retinopathy, green channel, hard exudates, image complements, optimization, thresholding",

author = "Hussain, \{Muhammad Altaf\} and Islam, \{Syed Osama Bin\} and Tiwana, \{M. I.\} and Ubaid-Ur-Rehman and Qureshi, \{W. S.\}",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 5th International Conference on Control, Automation and Robotics, ICCAR 2019 ; Conference date: 19-04-2019 Through 22-04-2019",

year = "2019",

month = apr,

doi = "10.1109/ICCAR.2019.8813469",

language = "English",

series = "2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019",

publisher = "IEEE",

pages = "206--211",

booktitle = "2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019",

}

Hussain, MA, Islam, SOB, Tiwana, MI, Ubaid-Ur-Rehman & Qureshi, WS 2019, Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks. in 2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019., 8813469, 2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019, IEEE, pp. 206-211, 5th International Conference on Control, Automation and Robotics, ICCAR 2019, Beijing, China, 19/04/19. https://doi.org/10.1109/ICCAR.2019.8813469

Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks. / Hussain, Muhammad Altaf; Islam, Syed Osama Bin; Tiwana, M. I. et al.
2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019. IEEE, 2019. p. 206-211 8813469 (2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019).

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

TY - GEN

T1 - Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks

AU - Hussain, Muhammad Altaf

AU - Islam, Syed Osama Bin

AU - Tiwana, M. I.

AU - Ubaid-Ur-Rehman,

AU - Qureshi, W. S.

PY - 2019/4

Y1 - 2019/4

N2 - Diabetic Retinopathy (DR) is an eye disorder that progressively leads to vision loss due to high glucose causing impairment of retinal blood vessels (BVs). 'Retinal Bright Lesions' such as 'Hard Exudates' (HEs) are plasma leakages from rapture retinal capillaries. HEs appear as hard, waxy, yellowish deposits from tiny spots to fat patches and signify moderate-severe Non-Proliferative Diabetic Retinopathy (NPDR). This paper proposes a simple, compact and computationally inexpensive technique for detection and classification of HEs using Digital Image Processing Techniques on digital fundus images complements and Artificial Neural Networks (ANN). The proposed technique unfolds through five stages i.e. Pre-processing, coarse detection, optimization, features detection extraction followed by classification. 'Speed Up Robust Features' (SURF) algorithm has been used for features detection extraction while 'Feed-Forward Back-propagation' (FFBP) ANN has been used for classification. The proposed technique has yielded 98.7% 'Sensitivity' (SE), 97.5% 'Specificity' (SP) and 97.7% 'Accuracy' (AC) on 'DIARETDB1' fundus images.

AB - Diabetic Retinopathy (DR) is an eye disorder that progressively leads to vision loss due to high glucose causing impairment of retinal blood vessels (BVs). 'Retinal Bright Lesions' such as 'Hard Exudates' (HEs) are plasma leakages from rapture retinal capillaries. HEs appear as hard, waxy, yellowish deposits from tiny spots to fat patches and signify moderate-severe Non-Proliferative Diabetic Retinopathy (NPDR). This paper proposes a simple, compact and computationally inexpensive technique for detection and classification of HEs using Digital Image Processing Techniques on digital fundus images complements and Artificial Neural Networks (ANN). The proposed technique unfolds through five stages i.e. Pre-processing, coarse detection, optimization, features detection extraction followed by classification. 'Speed Up Robust Features' (SURF) algorithm has been used for features detection extraction while 'Feed-Forward Back-propagation' (FFBP) ANN has been used for classification. The proposed technique has yielded 98.7% 'Sensitivity' (SE), 97.5% 'Specificity' (SP) and 97.7% 'Accuracy' (AC) on 'DIARETDB1' fundus images.

KW - artificial neural networks

KW - back-propagation

KW - confusion matrix

KW - diabetic retinopathy

KW - green channel

KW - hard exudates

KW - image complements

KW - optimization

KW - thresholding

UR - https://www.scopus.com/pages/publications/85071098638

U2 - 10.1109/ICCAR.2019.8813469

DO - 10.1109/ICCAR.2019.8813469

M3 - Conference contribution

AN - SCOPUS:85071098638

T3 - 2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019

SP - 206

EP - 211

BT - 2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019

PB - IEEE

T2 - 5th International Conference on Control, Automation and Robotics, ICCAR 2019

Y2 - 19 April 2019 through 22 April 2019

ER -

Hussain MA, Islam SOB, Tiwana MI, Ubaid-Ur-Rehman, Qureshi WS. Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks. In 2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019. IEEE. 2019. p. 206-211. 8813469. (2019 5th International Conference on Control, Automation and Robotics, ICCAR 2019). doi: 10.1109/ICCAR.2019.8813469

Detection and Classification of Hard Exudates with Fundus Images Complements and Neural Networks

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this