The Augmentation Data of Retina Image for Blood Vessel Segmentation Using U-Net Convolutional Neural Network Method

Erwin, Erwin (2022) The Augmentation Data of Retina Image for Blood Vessel Segmentation Using U-Net Convolutional Neural Network Method. International Journal of Computational Intelligence and Applications, 21 (1). pp. 1-17. ISSN 1757-5885

This is the latest version of this item.

[thumbnail of artikel IJCIA-S1469026822500043.pdf]
Preview
 Text
artikel IJCIA-S1469026822500043.pdf
Download (6MB) | Preview

Abstract

The retina is the most important part of the eye. By proper feature extraction, it can be the first step to detect a disease. Morphology of retina blood vessels can be used to identify and classify a disease. A step, such as segmentation and analysis of retinal blood vessels, can assist medical personnel in detecting the severity of a disease. In this paper, vascular segmentation using U-net architecture in the Convolutional Neural Network (CNN) method is proposed to train a sematic segmentation model in retinal blood vessel. In addition, the Contrast Limited Adaptive Histogram Equalization (CLAHE) method is used to increase the contrast of the grayscale and Median Filter is used to obtain better image quality. Data augmentation is also used to maximize the number of datasets owned to make more. The proposed method allows for easier implementation. In this study, the dataset used was STARE with the result of accuracy, sensitivity, specificity, precision, and F1-score that reached 97.64%, 78.18%, 99.20%, 88.77%, and 82.91%.

Item Type: Article
Subjects: Q Science > Q Science (General) > Q334-342 Computer science. Artificial intelligence. Algorithms. Robotics. Automation.
Divisions: 09-Faculty of Computer Science > 56201-Computer Systems (S1)
Depositing User: Dr Erwin Erwin
Date Deposited: 08 Aug 2022 07:26
Last Modified: 08 Aug 2022 07:26
URI: http://repository.unsri.ac.id/id/eprint/76525

Available Versions of this Item

Actions (login required)

View Item View Item