Related Experiment Video
Updated: Oct 23, 2025
05:58
Retinal Detachment Model in Rodents by Subretinal Injection of Sodium Hyaluronate
Published on: September 11, 2013
21.1K
RETRACTED ARTICLE: Region of interest-based predictive algorithm for subretinal hemorrhage detection using faster
M Suchetha1, N Sai Ganesh2, Rajiv Raman3
1Centre for Healthcare Advancement, Innovation and Research, VIT Chennai, Chennai, India.
Soft Computing
|August 23, 2021
Summary
This study introduces a deep learning algorithm using Faster R-CNN to detect Subretinal hemorrhage, improving diagnostic accuracy for macular edema (ME). The method efficiently classifies hemorrhage responsiveness, aiding in visual impairment treatment.
Area of Science:
- Ophthalmology
- Artificial Intelligence
- Medical Imaging
Background:
- Macular edema (ME) is a significant cause of vision loss, often linked to Age-related Macular Degeneration (AMD) and diabetic macular edema (DME).
- Early and accurate detection of complications like Subretinal hemorrhage is crucial for effective treatment and vision preservation.
Purpose of the Study:
- To develop and evaluate a deep learning-based predictive algorithm for detecting Subretinal hemorrhage.
- To improve the classification accuracy of responsive versus non-responsive Subretinal hemorrhage.
- To enhance the efficiency of diagnostic processes for macular edema.
Main Methods:
- Utilized Faster Region Convolutional Neural Network (Faster R-CNN) for detecting Subretinal hemorrhage in optical coherence tomography (OCT) images.
- Employed semantic segmentation to isolate the Region of Interest (ROI) containing the hemorrhage.
- Trained the algorithm on a dataset from a medical institution and validated using a Kaggle dataset for comparison with traditional Convolutional Neural Network (CNN) methods.
Main Results:
- Achieved an average sensitivity of 85.3%, selectivity of 89.64%, and accuracy of 93.48% on combined datasets.
- Demonstrated a lower time complexity of 2.64s for testing compared to traditional CNN, R-CNN, and Fast R-CNN methods.
- The algorithm successfully detected Subretinal hemorrhage and classified its responsiveness.
Conclusions:
- The proposed Faster R-CNN based deep learning algorithm offers a highly accurate and efficient method for detecting Subretinal hemorrhage.
- This approach has the potential to significantly aid in the diagnosis and management of macular edema, contributing to better patient outcomes.
- The algorithm's speed and accuracy make it a promising tool for clinical application in ophthalmology.
