Technology Category
Machine Learning: Transfer, Domain Adaptation, Multi-Task LearningComputer Vision: Learning & Optimization for CVSearch and Optimization: Learning to Search
Early detection of diabetes through transfer learning-based eye (vision) screening and improvement of machine learning model performance and advanced parameter setting algorithms
🤖 AI Summary
To address the low accuracy, prolonged training time, and poor generalization of conventional CNNs in early diabetic retinopathy (DR) screening—particularly under small-sample conditions—this paper proposes a novel transfer learning framework integrating dimensionality reduction, dynamic learning rate adaptation, and Bayesian hyperparameter optimization. Building upon a pre-trained CNN backbone, the framework incorporates feature dimensionality reduction to mitigate overfitting, employs a dynamically adjusted learning rate schedule to accelerate convergence, and leverages Bayesian optimization for efficient, global search of optimal hyperparameters. Evaluated on standard fundus image datasets, the model achieves an overall test accuracy of 84% (with a peak class-wise accuracy of 89%), sensitivity of 97%, and an F1-score of 92%, substantially outperforming existing approaches. This work establishes a reusable, robust transfer learning paradigm tailored for small-sample medical image diagnosis.
Technology Category
Application Category
📝 Abstract
Diabetic Retinopathy (DR) is a serious and common complication of diabetes, caused by prolonged high blood sugar levels that damage the small retinal blood vessels. If left untreated, DR can progress to retinal vein occlusion and stimulate abnormal blood vessel growth, significantly increasing the risk of blindness. Traditional diabetes diagnosis methods often utilize convolutional neural networks (CNNs) to extract visual features from retinal images, followed by classification algorithms such as decision trees and k-nearest neighbors (KNN) for disease detection. However, these approaches face several challenges, including low accuracy and sensitivity, lengthy machine learning (ML) model training due to high data complexity and volume, and the use of limited datasets for testing and evaluation. This study investigates the application of transfer learning (TL) to enhance ML model performance in DR detection. Key improvements include dimensionality reduction, optimized learning rate adjustments, and advanced parameter tuning algorithms, aimed at increasing efficiency and diagnostic accuracy. The proposed model achieved an overall accuracy of 84% on the testing dataset, outperforming prior studies. The highest class-specific accuracy reached 89%, with a maximum sensitivity of 97% and an F1-score of 92%, demonstrating strong performance in identifying DR cases. These findings suggest that TL-based DR screening is a promising approach for early diagnosis, enabling timely interventions to prevent vision loss and improve patient outcomes.
Problem
Research questions and friction points this paper is trying to address.
Improving diabetes detection via transfer learning-based eye screening
Enhancing ML model accuracy and sensitivity for Diabetic Retinopathy
Optimizing parameter settings to reduce training time and complexity
Innovation
Methods, ideas, or system contributions that make the work stand out.
Transfer learning enhances diabetic retinopathy detection
Advanced parameter tuning boosts model accuracy
Dimensionality reduction optimizes learning efficiency
🔎 Similar Papers
No similar papers found.
