Automatic Diagnosis of Familial Exudative Vitreoretinopathy Using a Fusion Neural Network for Wide-Angle Retinal Images
Open Access
- 23 December 2019
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 162-173
Abstract
Familial exudative vitreoretinopathy (FEVR) is a hereditary disorder that can damage the retina. This retinal damage can lead to vision loss and even blindness in the late stages. Thus, early diagnosis and prevention of the disease’s progression are critical. The purpose of this study was to develop an automated diagnosis system for FEVR based on combining deep learning and domain knowledge. A transfer learning scheme was designed to train a deep convolutional neural network (DCNN) to provide segmentation of the retinal vessels. Based on this vessel segmentation and prior clinical knowledge, the vascular characteristics, including the retinal avascular area, vessel angle, fractal dimension, branching and density of blood vessels, were automatically evaluated. Finally, the diagnosis of FEVR was achieved by a feature fusion neural network. Our method was evaluated on 300 images with 168 healthy and 132 FEVR images. By combining deep features and handcrafted features (extracted vascular characteristics), the proposed method achieved an average F1-score of 0.95, with excellent accuracy (94.34%) and sensitivity (91.43%); the quadratic weighted κ was 0.88 for the diagnosis of FEVR. We demonstrated the effectiveness and robustness of the proposed method using five-fold cross-validation. The proposed automatic diagnosis system can assist doctors for better judgment and make sense of early diagnosis and prevention of the disease’s progression.Keywords
Funding Information
- National Natural Science Foundation of China (61836011)
- Fundamental Research Funds for the Central Universities (WK2090090023)
This publication has 31 references indexed in Scilit:
- An Image Stitching Algorithm Based on Histogram Matching and SIFT AlgorithmInternational Journal of Pattern Recognition and Artificial Intelligence, 2017
- The Morphology of the Brain Neurons: Box-Counting Method in Quantitative Analysis of 2D Image2016
- Ultra-wide-field scanning laser ophthalmoscopy assists in the clinical detection and evaluation of asymptomatic early-stage familial exudative vitreoretinopathyGraefe's Archive for Clinical and Experimental Ophthalmology, 2016
- ULTRA-WIDEFIELD FUNDUS IMAGINGRetina, 2016
- Deep Dynamic Neural Networks for Multimodal Gesture Segmentation and RecognitionIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Fractal Analysis of Two Dimensional Images: Parameters of the Space-Filling and Shape2015
- Surgical Outcomes of Progressive Tractional Retinal Detachment Associated With Familial Exudative VitreoretinopathyAmerican Journal of Ophthalmology, 2014
- Diversity of Retinal Vascular Anomalies in Patients with Familial Exudative VitreoretinopathyOphthalmology, 2014
- The role of Frizzled-4 mutations in familial exudative vitreoretinopathy and Coats diseaseBritish Journal of Ophthalmology, 2010
- Familial Exudative VitreoretinopathyAmerican Journal of Ophthalmology, 1969