Research on Diabetic Retinopathy Auxiliary Diagnostic System Combining Deep Learning and Medical Big Data

DOI: https://doi.org/10.62517/jes.202602107

Author(s)

Liu Jiaxin, Long Yanbin*

Affiliation(s)

Liaoning University of Science and Technology, Anshan, China *Corresponding Author

Abstract

Diabetic retinopathy (DR), as the most common and blinding microvascular complication of diabetes, requires early screening and accurate diagnosis to prevent vision loss. Traditional diagnosis relies on ophthalmologists manually interpreting fundus images, which suffers from low efficiency, high subjectivity, and uneven resource distribution. With breakthroughs in deep learning technology and the accumulation of medical big data, deep learning-based DR-assisted diagnostic systems have shown revolutionary potential. This paper systematically reviews the technical path of deep learning in DR diagnosis, including data preprocessing, model architecture design, multimodal data fusion, and clinical validation methods. It analyzes its advantages in improving diagnostic efficiency, reducing missed diagnosis rates, and optimizing the allocation of medical resources, and discusses the challenges in system deployment and future development directions. Research shows that the combination of deep learning and medical big data provides an innovative solution for accurate screening and personalized treatment of DR, with broad clinical application prospects.

Keywords

Deep Learning; Medical Big Data; Diabetic Retinopathy; Auxiliary Diagnostic System; Convolutional Neural Network; Multimodal Fusion

References

[1] Wu Yongqi; Xiao Yanshan; Liu Bo. Image classification based on ordered regression and dictionary learning. Computer Applications and Software, 2025(10)76-77 [2] Qi Xiangming; Liu Xiaowei. Image classification network enhanced by dual wavelet. Pattern Recognition and Artificial Intelligence, 2025(09)56-58 [3] Song Guiping. Performance breakthrough and practice of meta-learning to assist small sample image classification. Communication World, 2025(09)45-47 [4] Wu Huiling; Zhou Pingping. Flower image classification based on feature mining and cross-layer complementary learning. Modern Information Technology, 2025(17)33-34 [5] Liu Shaoyu; Zhang Zijun; Zhou Zhidan. Research on image classification method based on diffusion model. Computer Technology and Development, 2025(12)78-79 [6] Li Shengwei; Li Hai; Mark; Zhao Hongwei. Image classification framework based on self-distillation. Information Technology, 2025(04)55-56 [7] Cai Yuhong; Qiao Lina; Xiao Jiji; Chen Zijie. Research on training and evaluation of image classification model based on transfer learning. Electronic Quality, 2025(04)98-99 [8] Bai Ruifeng; Gou Guanglei; Wen Lang; Miao Wanyu. Small sample image classification method based on particle-sphere prototype network. Computer Applications, 2025(07)67-69 [9] Shen Yanping; Han Shaoyong; Gu Suhang; Xun Zhan. Robust multi-view latent low-rank representation image classification method. Journal of Shihezi University (Natural Science Edition), 2024(05)2-3 [10] Yang Chuangguang; Chen Luming; Zhao Erhu; An Zhulin; Xu Yongjun. Image classification method based on graph representation knowledge distillation. Acta Electronica Sinica, 2024(10)45-48