Abstract
This paper presents a retinal vessel segmentation method that is inspired by the
human visual system and uses a Gabor filter bank. Machine learning is used to
optimize the filter parameters for retinal vessel extraction. The filter responses are represented as textons and this allows the corresponding membership functions to be used as the framework for learning vessel and non-vessel classes. Then, vessel texton memberships are used to generate segmentation results. We evaluate our method using the publicly available DRIVE database. It achieves competitive performance (sensitivity=0.7673, specificity=0.9602, accuracy=0.9430) compared to other recently published work. These figures are particularly interesting as our filter bank is quite generic and only includes Gabor responses. Our experimental results also show that the performance, in terms of sensitivity, is superior to other methods.
human visual system and uses a Gabor filter bank. Machine learning is used to
optimize the filter parameters for retinal vessel extraction. The filter responses are represented as textons and this allows the corresponding membership functions to be used as the framework for learning vessel and non-vessel classes. Then, vessel texton memberships are used to generate segmentation results. We evaluate our method using the publicly available DRIVE database. It achieves competitive performance (sensitivity=0.7673, specificity=0.9602, accuracy=0.9430) compared to other recently published work. These figures are particularly interesting as our filter bank is quite generic and only includes Gabor responses. Our experimental results also show that the performance, in terms of sensitivity, is superior to other methods.
Original language | English |
---|---|
Pages | 155-160 |
Number of pages | 6 |
Publication status | Published - 2014 |
Event | Medical Image Understanding and Analysis (MIUA 2014) - Royal Holloway College, London, United Kingdom Duration: 9 Jul 2014 → 11 Jul 2014 |
Conference
Conference | Medical Image Understanding and Analysis (MIUA 2014) |
---|---|
Country/Territory | United Kingdom |
City | London |
Period | 9/07/14 → 11/07/14 |
Profiles
-
Wenjia Wang
- School of Computing Sciences - Professor of Artificial Intelligence
- Data Science and AI - Member
Person: Research Group Member, Academic, Teaching & Research