Related Experiment Video
Updated: Jul 16, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.8K
Protocol to analyze fundus images for multidimensional quality grading and real-time guidance using deep learning
Lixue Liu1, Mingyuan Li1, Duoru Lin1
1State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangdong Provincial Key Laboratory of Ophthalmology and Vision Science, Guangdong Provincial Clinical Research Center for Ocular Diseases, Guangzhou, Guangdong, China.
STAR Protocols
|September 21, 2023
Summary
DeepFundus uses deep learning to classify fundus image quality, improving medical AI research by guiding real-time image acquisition and ensuring high-quality data.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Data quality is a major challenge in medical artificial intelligence (AI) research.
- High-quality fundus images are crucial for accurate diagnosis and AI model training.
Purpose of the Study:
- To introduce DeepFundus, a deep learning system for multidimensional fundus image quality classification.
- To provide real-time, on-site guidance for acquiring high-quality fundus images.
Main Methods:
- Utilized deep learning techniques for image quality assessment.
- Developed a protocol for data preparation, model training, inference, and evaluation.
- Implemented results visualization using heatmaps.
- Protocol designed for Python implementation with customizable datasets.
Main Results:
- DeepFundus enables multidimensional classification of fundus image quality.
- The system offers real-time feedback for improving image acquisition.
- Heatmaps visualize model performance and areas of interest.
Conclusions:
- DeepFundus addresses critical data quality issues in medical AI.
- The protocol facilitates the development and deployment of robust fundus image analysis tools.
- This approach can enhance the reliability of AI in ophthalmology.
