Xing WangMeiYu, Li XiaoNa, Ni JingShu, Zhang YuanZhi, Li ZhongSheng, Liu Yong, Wang YiKun, Huang Yao
College of Biomedical Engineering, Anhui Medical University, Hefei, 230011, China.
Anhui Institute of Optics and Fine Mechanics, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Anhui Provincial Engineering Technology Research Center for Biomedical Optical Instrument, Anhui Provincial Engineering Laboratory for Medical Optical Diagnosis Treatment Technology and Instrument, Hefei, 230031, China.
Biomed Eng Online. 2025 Jul 9;24(1):85. doi: 10.1186/s12938-025-01416-2.
Myopia is a major cause of vision impairment. To improve the efficiency of myopia screening, this paper proposes a deep learning model, X-ENet, which combines the advantages of depthwise separable convolution and dynamic convolution to classify different severities of myopia. The proposed model not only enables precise extraction of detailed features from fundus images but also achieves lightweight processing, thereby improving both computational efficiency and classification accuracy.
First, fundus images are enhanced and preprocessed to improve feature extraction effectiveness and enhance the model's generalization capability. Then, the model is trained using fivefold cross-validation, leveraging dynamic convolution and depthwise separable convolution to extract features from each fundus image and classify the severity of myopia. Next, Grad-CAM is employed to visualize the model's decision-making process, highlighting the regions contributing to classification. Finally, a user-friendly GUI interface is developed to intuitively present the classification results, thereby enhancing the system's usability and practical applicability.
The experimental results show that the proposed method achieves an accuracy of 0.9104, a precision of 0.8154, a recall of 0.8177, an F1-score of 0.8147, and a specificity of 0.9376 in the classification of myopia severity.
The model significantly outperforms existing conventional deep learning models in terms of accuracy, demonstrating strong effectiveness and reliability.
近视是视力损害的主要原因。为提高近视筛查效率，本文提出一种深度学习模型X-ENet，它结合了深度可分离卷积和动态卷积的优点来对不同严重程度的近视进行分类。所提出的模型不仅能够从眼底图像中精确提取详细特征，还能实现轻量级处理，从而提高计算效率和分类准确率。
首先，对眼底图像进行增强和预处理，以提高特征提取效果并增强模型的泛化能力。然后，使用五折交叉验证对模型进行训练，利用动态卷积和深度可分离卷积从每个眼底图像中提取特征并对近视严重程度进行分类。接下来，采用Grad-CAM来可视化模型的决策过程，突出对分类有贡献的区域。最后，开发一个用户友好的GUI界面来直观呈现分类结果，从而提高系统的可用性和实际适用性。
实验结果表明，所提出的方法在近视严重程度分类中达到了0.9104的准确率、0.8154的精确率、0.8177的召回率、0.8147的F1分数和0.9376的特异性。
该模型在准确率方面显著优于现有的传统深度学习模型，显示出强大的有效性和可靠性。
