An Insight on the Timely Diagnosis of Diabetic Retinopathy Using Traditional and AI-Driven Approaches
Diabetic Retinopathy is a progressive microvascular complication of diabetes that requires early detection to improve patient outcomes. Traditional screening techniques, including fundus photography and optical coherence tomography, provide valuable diagnostic insights but have some limitations, lik...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11053490/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Diabetic Retinopathy is a progressive microvascular complication of diabetes that requires early detection to improve patient outcomes. Traditional screening techniques, including fundus photography and optical coherence tomography, provide valuable diagnostic insights but have some limitations, like cost and technical complexity. Artificial intelligence is transforming the detection of diabetic retinopathy, moving away from traditional machine learning models that rely on manually created features to deep learning methods that allow for automatic feature extraction from retinal images. This systematic review investigates the evolution and diagnostic performance of AI-based techniques for DR detection. Studies were included if they applied machine learning or deep learning methods to retinal fundus or OCT images for DR classification. A total of 116 studies were included following comprehensive searches in databases such as PubMed, ScienceDirect, and IEEE Xplore, covering publications up to February 2024. Risk of bias was assessed in a representative sample of six studies, indicating a significant overall risk due to inadequate reporting on blinding and selective outcome reporting. Federated learning emerged as a promising alternative, enabling decentralized collaboration without compromising data privacy. Additionally, the growing focus on Explainable AI helps address the “non-explainable” nature of deep learning models by providing visual and textual explanations for predictions, thereby enhancing clinician trust and facilitating informed decision-making. By incorporating artificial intelligence with standard diagnostic frameworks, this research highlights the possibility for more accurate, scalable, and reachable diabetic retinopathy detection, paving the way for considerable advancements in ophthalmic problem management and enhancing patient care. |
|---|---|
| ISSN: | 2169-3536 |