Title:
Hybrid Edr-Net+ Architecture with Noise Invariant and Effective Computation for Alternative Classification of Referable Diabetic Retinopathy

Author:

Supervisor:

Department:

Electrical & Electronics Engineering

Year:

2024

URL:

Abstract:

The global prevalence of Diabetic Retinopathy (DR) has heightened the urgency to develop an AI-driven automated referable DR (rDR) screening system. Consequently, there’s a rising focus on researchi...
View more

Multimedia Content:

01

Methodology

The methodology used in this thesis focuses on simulation-based analysis. The main steps are: 1. Simulation Environment Built using MATLAB as the main platform. Simulations modeled Mobile Crowdse...
View more
Methodology Image
02

Citation Analysis

1. IEEE Transactions on Neural Networks and Learning Systems Article Example: Deep learning architectures for medical image classification. Scope: Neural networks, learning systems, and their appl...
View more
03

Impact

1. Societal Impact Encourages more people to participate in MCS projects by giving fair and motivating rewards. Leads to better community-driven applications like smart traffic monitoring, environ...
View more
04

Benefits to Society

1. Improved Participation The hybrid incentive (money + reputation) motivates more users to join and stay active in MCS projects. 2. Higher Data Quality By rewarding reputation as well as mon...
View more
05

Applications

1. Mobile Crowdsensing (MCS) Applications The thesis directly applies to MCS platforms, where smartphone users act as sensors (e.g., traffic monitoring, environmental sensing, public health trackin...
View more
06

Infographic

Summary in infographic
Infographic Image
07

Patents / Products

1. Hybrid Incentive Management Engine •Product Type: Software module or API •Use Case: Integrate into Mobile Crowdsensing (MCS) platforms or smart city apps to balance monetary and reputation rew...
View more
08

Experts

1. Prof. Dr. Tang Tong Boon Contribution: Supervised the development of a novel deep learning architecture (EDR-Net+) for diabetic retinopathy detection. Key Work: Guided research on hybrid convol...
View more