Enhancement of Van Herick's method for automatic diagnosis of angle-closure glaucoma

Name: Ardakani, Ali Sadeghi

keyword: Angle-closure glaucoma

; Dimensionality reduction

; Gaussian mixture models

; Convolutional deep neural networks

Abstract: According to ophthalmological speculations, Angle Closure Glaucoma (ACG) is going to be a major risk to Asian demography’s vision in the coming years. Therefore, devising plans and infrastructures to encounter this hazardous sight disease gained higher priorities. One of the challenges, is the ability to provide the screening and diagnosis facilities for remote, and unprivileged areas. As a solution, a relatively portable platform of automatic ACG screening is designed and developed by National Electronics and Computer Technology Center (NECTEC), a member of National Science and Technology Development Agency (NSTDA). This platform’s main objective is to perform ACG diagnosis without reliance on professionally trained ophthalmologist. Previously, various methods introduced to perform the Van Herick’s methods measurement and diagnosis automatically on selected set of slit-lamp images obtained by professional ophthalmologist. Indeed, there was involvement of professional operator remaining as initial part of procedure which in part contributed to inconsistencies in results. This study, investigated the highlighted role and provided methods to successfully eliminate inner loop involvement of professional operator in VH’s method based automatic screening of ACG. Furthermore, it reduces methodological involvement of the experts in the process by providing automatic feature selection methods and trainable convolutional layers. In addition, utilization of powerful classifiers like Deep Neural Networks, AdaBoost, and Random Forests on top resulted in largely satisfactory outcomes in the end. Meanwhile, this study suggests more robust means of data collection in order to strengthen the upshot of the diagnosis

Thammasat University. Thammasat University Library

Address: BANGKOK

Email: preserv@tu.ac.th

Name: Waree Kongprawechnon

Role: advisor

Name: Thanaruk Theeramunkong

Role: co-advisor

Created: 2016

Modified: 2021-06-23

Issued: 2021-06-23

วิทยานิพนธ์/Thesis

application/pdf

DOI: 10.14457/TU.the.2016.1765

eng

DegreeName: Master of Engineering

Level: Master's Degree

Descipline: Information and Communication Technology for Embedded Systems

Grantor: Thammasat University

©copyrights Thammasat University

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	2961ardakani.pdf	15.36 MB	12	2026-05-26 18:16:01

ใช้เวลา

0.033056 วินาที