Abstract
Screening for vision threatening diabetic retinopathy by grading digital retinal images reduces the risk of blindness in people with diabetes. Computer-aided diagnosis can aid human graders to cope with this mounting problem. We propose to use a 10-layer convolutional neural network to automatically, simultaneously segment and discriminate exudates, haemorrhages and micro-aneurysms. Input image is normalized before segmentation. The net is trained in two stages to improve performance. On average, our net on 30,275,903 effective points achieved a sensitivity of 0.8758 and 0.7158 for exudates and dark lesions on the CLEOPATRA database. It also achieved a sensitivity of 0.6257 and 0.4606 for haemorrhages and micro-aneurysms. This study shows that it is possible to get a single convolutional neural network to segment these pathological features on a wide range of fundus images with reasonable accuracy.
Original language | English |
---|---|
Pages (from-to) | 66-76 |
Number of pages | 11 |
Journal | Information Sciences |
Volume | 420 |
DOIs | |
Publication status | Published - 01-12-2017 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Software
- Control and Systems Engineering
- Theoretical Computer Science
- Computer Science Applications
- Information Systems and Management
- Artificial Intelligence
Cite this
}
Automated segmentation of exudates, haemorrhages, microaneurysms using single convolutional neural network. / Tan, Jen Hong; Fujita, Hamido; Sivaprasad, Sobha; Bhandary, Sulatha V.; Rao, A. Krishna; Chua, Kuang Chua; Acharya, U. Rajendra.
In: Information Sciences, Vol. 420, 01.12.2017, p. 66-76.Research output: Contribution to journal › Article
TY - JOUR
T1 - Automated segmentation of exudates, haemorrhages, microaneurysms using single convolutional neural network
AU - Tan, Jen Hong
AU - Fujita, Hamido
AU - Sivaprasad, Sobha
AU - Bhandary, Sulatha V.
AU - Rao, A. Krishna
AU - Chua, Kuang Chua
AU - Acharya, U. Rajendra
PY - 2017/12/1
Y1 - 2017/12/1
N2 - Screening for vision threatening diabetic retinopathy by grading digital retinal images reduces the risk of blindness in people with diabetes. Computer-aided diagnosis can aid human graders to cope with this mounting problem. We propose to use a 10-layer convolutional neural network to automatically, simultaneously segment and discriminate exudates, haemorrhages and micro-aneurysms. Input image is normalized before segmentation. The net is trained in two stages to improve performance. On average, our net on 30,275,903 effective points achieved a sensitivity of 0.8758 and 0.7158 for exudates and dark lesions on the CLEOPATRA database. It also achieved a sensitivity of 0.6257 and 0.4606 for haemorrhages and micro-aneurysms. This study shows that it is possible to get a single convolutional neural network to segment these pathological features on a wide range of fundus images with reasonable accuracy.
AB - Screening for vision threatening diabetic retinopathy by grading digital retinal images reduces the risk of blindness in people with diabetes. Computer-aided diagnosis can aid human graders to cope with this mounting problem. We propose to use a 10-layer convolutional neural network to automatically, simultaneously segment and discriminate exudates, haemorrhages and micro-aneurysms. Input image is normalized before segmentation. The net is trained in two stages to improve performance. On average, our net on 30,275,903 effective points achieved a sensitivity of 0.8758 and 0.7158 for exudates and dark lesions on the CLEOPATRA database. It also achieved a sensitivity of 0.6257 and 0.4606 for haemorrhages and micro-aneurysms. This study shows that it is possible to get a single convolutional neural network to segment these pathological features on a wide range of fundus images with reasonable accuracy.
UR - http://www.scopus.com/inward/record.url?scp=85027565220&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85027565220&partnerID=8YFLogxK
U2 - 10.1016/j.ins.2017.08.050
DO - 10.1016/j.ins.2017.08.050
M3 - Article
AN - SCOPUS:85027565220
VL - 420
SP - 66
EP - 76
JO - Information Sciences
JF - Information Sciences
SN - 0020-0255
ER -