Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss

Usman Naseem; Matloob Khushi; Shah Khalid Khan; Nazar Waheed; Adnan Mir; Atika Qazi; Bandar Alshammari; Simon K. Poon

doi:10.1007/978-3-030-63836-8_3

Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss

Usman Naseem^*, Matloob Khushi, Shah Khalid Khan, Nazar Waheed, Adnan Mir, Atika Qazi, Bandar Alshammari, Simon K. Poon

^*Corresponding author for this work

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

11 Citations (Scopus)

Abstract

Diabetic retinopathy (DR) is the most common eye threatening micro-vascular complication of diabetes. It develops and grows without arbitrary symptoms and can ultimately lead to blindness. However, 90% of the DR-attributed blindness is preventable but needs prompt diagnosis and appropriate treatment. Presently, DR detection is time and resource-consuming, i.e., required qualified ophthalmologist technician to examine the retina colour fundus for investigating the existence of vascular anomaly associated lesions. Nevertheless, an automatic DR scanning with specialised deep learning algorithms can overcome this challenge. In this paper, we present an automatic detection of DR using Multi-layer Neural Networks and Split Attention with Focal Loss. Our method outperformed state-of-the-art (SOTA) networks in early-stage detection and achieved 85.9% accuracy in DR classification. Because of high performance, it is believed that the results obtained in this paper are of great importance to the medical and the relevant research community.

Original language	English
Title of host publication	Neural Information Processing
Subtitle of host publication	27th International Conference, ICONIP 2020, Bangkok, Thailand, November 23–27, 2020, proceedings, part III
Editors	Haiqin Yang, Kitsuchart Pasupa, Andrew Chi-Sing Leung, James T. Kwok, Jonathan H. Chan, Irwin King
Place of Publication	Cham
Publisher	Springer, Springer Nature
Pages	26-37
Number of pages	12
ISBN (Electronic)	9783030638368
ISBN (Print)	9783030638351
DOIs	https://doi.org/10.1007/978-3-030-63836-8_3
Publication status	Published - 2020
Externally published	Yes
Event	27th International Conference on Neural Information Processing, ICONIP 2020 - Bangkok, Thailand Duration: 18 Nov 2020 → 22 Nov 2020

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	12534
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	27th International Conference on Neural Information Processing, ICONIP 2020
Country/Territory	Thailand
City	Bangkok
Period	18/11/20 → 22/11/20

Keywords

Diabetic Retinopathy
Deep learning
Computer-aided diagnosis
ResNet
Split attention
Ophthalmoscopy

Access to Document

10.1007/978-3-030-63836-8_3

Cite this

Naseem, U., Khushi, M., Khan, S. K., Waheed, N., Mir, A., Qazi, A., Alshammari, B., & Poon, S. K. (2020). Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss. In H. Yang, K. Pasupa, A. C.-S. Leung, J. T. Kwok, J. H. Chan, & I. King (Eds.), Neural Information Processing: 27th International Conference, ICONIP 2020, Bangkok, Thailand, November 23–27, 2020, proceedings, part III (pp. 26-37). (Lecture Notes in Computer Science; Vol. 12534). Springer, Springer Nature. https://doi.org/10.1007/978-3-030-63836-8_3

Naseem, Usman ; Khushi, Matloob ; Khan, Shah Khalid et al. / Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss. Neural Information Processing: 27th International Conference, ICONIP 2020, Bangkok, Thailand, November 23–27, 2020, proceedings, part III. editor / Haiqin Yang ; Kitsuchart Pasupa ; Andrew Chi-Sing Leung ; James T. Kwok ; Jonathan H. Chan ; Irwin King. Cham : Springer, Springer Nature, 2020. pp. 26-37 (Lecture Notes in Computer Science).

@inproceedings{59403a7a3612419e84ea01cd0bde14b1,

title = "Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss",

abstract = "Diabetic retinopathy (DR) is the most common eye threatening micro-vascular complication of diabetes. It develops and grows without arbitrary symptoms and can ultimately lead to blindness. However, 90\% of the DR-attributed blindness is preventable but needs prompt diagnosis and appropriate treatment. Presently, DR detection is time and resource-consuming, i.e., required qualified ophthalmologist technician to examine the retina colour fundus for investigating the existence of vascular anomaly associated lesions. Nevertheless, an automatic DR scanning with specialised deep learning algorithms can overcome this challenge. In this paper, we present an automatic detection of DR using Multi-layer Neural Networks and Split Attention with Focal Loss. Our method outperformed state-of-the-art (SOTA) networks in early-stage detection and achieved 85.9\% accuracy in DR classification. Because of high performance, it is believed that the results obtained in this paper are of great importance to the medical and the relevant research community.",

keywords = "Diabetic Retinopathy, Deep learning, Computer-aided diagnosis, ResNet, Split attention, Ophthalmoscopy",

author = "Usman Naseem and Matloob Khushi and Khan, \{Shah Khalid\} and Nazar Waheed and Adnan Mir and Atika Qazi and Bandar Alshammari and Poon, \{Simon K.\}",

year = "2020",

doi = "10.1007/978-3-030-63836-8\_3",

language = "English",

isbn = "9783030638351",

series = "Lecture Notes in Computer Science",

publisher = "Springer, Springer Nature",

pages = "26--37",

editor = "Haiqin Yang and Kitsuchart Pasupa and Leung, \{Andrew Chi-Sing\} and Kwok, \{James T.\} and Chan, \{Jonathan H.\} and Irwin King",

booktitle = "Neural Information Processing",

address = "United States",

note = "27th International Conference on Neural Information Processing, ICONIP 2020 ; Conference date: 18-11-2020 Through 22-11-2020",

}

Naseem, U, Khushi, M, Khan, SK, Waheed, N, Mir, A, Qazi, A, Alshammari, B & Poon, SK 2020, Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss. in H Yang, K Pasupa, AC-S Leung, JT Kwok, JH Chan & I King (eds), Neural Information Processing: 27th International Conference, ICONIP 2020, Bangkok, Thailand, November 23–27, 2020, proceedings, part III. Lecture Notes in Computer Science, vol. 12534, Springer, Springer Nature, Cham, pp. 26-37, 27th International Conference on Neural Information Processing, ICONIP 2020, Bangkok, Thailand, 18/11/20. https://doi.org/10.1007/978-3-030-63836-8_3

Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss. / Naseem, Usman; Khushi, Matloob; Khan, Shah Khalid et al.
Neural Information Processing: 27th International Conference, ICONIP 2020, Bangkok, Thailand, November 23–27, 2020, proceedings, part III. ed. / Haiqin Yang; Kitsuchart Pasupa; Andrew Chi-Sing Leung; James T. Kwok; Jonathan H. Chan; Irwin King. Cham: Springer, Springer Nature, 2020. p. 26-37 (Lecture Notes in Computer Science; Vol. 12534).

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

TY - GEN

T1 - Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss

AU - Naseem, Usman

AU - Khushi, Matloob

AU - Khan, Shah Khalid

AU - Waheed, Nazar

AU - Mir, Adnan

AU - Qazi, Atika

AU - Alshammari, Bandar

AU - Poon, Simon K.

PY - 2020

Y1 - 2020

N2 - Diabetic retinopathy (DR) is the most common eye threatening micro-vascular complication of diabetes. It develops and grows without arbitrary symptoms and can ultimately lead to blindness. However, 90% of the DR-attributed blindness is preventable but needs prompt diagnosis and appropriate treatment. Presently, DR detection is time and resource-consuming, i.e., required qualified ophthalmologist technician to examine the retina colour fundus for investigating the existence of vascular anomaly associated lesions. Nevertheless, an automatic DR scanning with specialised deep learning algorithms can overcome this challenge. In this paper, we present an automatic detection of DR using Multi-layer Neural Networks and Split Attention with Focal Loss. Our method outperformed state-of-the-art (SOTA) networks in early-stage detection and achieved 85.9% accuracy in DR classification. Because of high performance, it is believed that the results obtained in this paper are of great importance to the medical and the relevant research community.

AB - Diabetic retinopathy (DR) is the most common eye threatening micro-vascular complication of diabetes. It develops and grows without arbitrary symptoms and can ultimately lead to blindness. However, 90% of the DR-attributed blindness is preventable but needs prompt diagnosis and appropriate treatment. Presently, DR detection is time and resource-consuming, i.e., required qualified ophthalmologist technician to examine the retina colour fundus for investigating the existence of vascular anomaly associated lesions. Nevertheless, an automatic DR scanning with specialised deep learning algorithms can overcome this challenge. In this paper, we present an automatic detection of DR using Multi-layer Neural Networks and Split Attention with Focal Loss. Our method outperformed state-of-the-art (SOTA) networks in early-stage detection and achieved 85.9% accuracy in DR classification. Because of high performance, it is believed that the results obtained in this paper are of great importance to the medical and the relevant research community.

KW - Diabetic Retinopathy

KW - Deep learning

KW - Computer-aided diagnosis

KW - ResNet

KW - Split attention

KW - Ophthalmoscopy

UR - https://www.scopus.com/pages/publications/85097419719

U2 - 10.1007/978-3-030-63836-8_3

DO - 10.1007/978-3-030-63836-8_3

M3 - Conference proceeding contribution

SN - 9783030638351

T3 - Lecture Notes in Computer Science

SP - 26

EP - 37

BT - Neural Information Processing

A2 - Yang, Haiqin

A2 - Pasupa, Kitsuchart

A2 - Leung, Andrew Chi-Sing

A2 - Kwok, James T.

A2 - Chan, Jonathan H.

A2 - King, Irwin

PB - Springer, Springer Nature

CY - Cham

T2 - 27th International Conference on Neural Information Processing, ICONIP 2020

Y2 - 18 November 2020 through 22 November 2020

ER -

Naseem U, Khushi M, Khan SK, Waheed N, Mir A, Qazi A et al. Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss. In Yang H, Pasupa K, Leung ACS, Kwok JT, Chan JH, King I, editors, Neural Information Processing: 27th International Conference, ICONIP 2020, Bangkok, Thailand, November 23–27, 2020, proceedings, part III. Cham: Springer, Springer Nature. 2020. p. 26-37. (Lecture Notes in Computer Science). doi: 10.1007/978-3-030-63836-8_3

Diabetic retinopathy detection using multi-layer neural networks and split attention with focal loss

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this