Glyconet image lesion segmentation identification method and system based on deep learning
A deep learning and lesion technology, applied in the field of sugar network image lesion segmentation and identification, can solve the problems of insufficient fine analysis performance of fundus sugar network image lesions, low timeliness accuracy of deep learning models, and performance limitations of processing equipment.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0019] like figure 1 As shown, the present application provides a deep learning-based method for segmentation and identification of sugar network image lesions, the method is applied to an intelligent lesion segmentation and identification system, and the intelligent lesion segmentation and identification system is connected in communication with the basic network set training module, and the Methods include:
[0020] S100: Call a pre-stored fundus image data set by the intelligent lesion segmentation and identification system, wherein the fundus image data set includes hemorrhage, microvascular tumor, hard exudation, soft exudation, neovascular fibrous proliferation membrane, and abnormal microvascular sugar reticulum lesions;
[0021] Specifically, the earliest lesions of diabetic retinopathy include microaneurysms and small hemorrhages, after which vascular changes can progress to capillary nonperfusion, resulting in hemorrhage, vitreous abnormalities, and intraretinal mic...
Embodiment approach
[0025] Optionally, as figure 2 As shown, an implementation manner of step S200 in the method provided in this embodiment of the present application includes:
[0026] S210: Build a generative model, wherein the generative model is a model for image generation;
[0027] S220: Input the fundus image dataset and random noise into the generation model to obtain a pre-expanded fundus image dataset;
[0028] S230: Identify the pre-expanded fundus image data set, and obtain expanded data and feedback data according to the identification result;
[0029] S240: adding the expanded data to the expanded fundus image dataset, feeding back the feedback data to the generation model for model optimization, and using the optimized generation model to generate a new expanded fundus image dataset;
[0030] S250: Repeat the process of feedback data and expanded data feedback and expansion according to the new expanded fundus image dataset to obtain the expanded fundus image dataset.
[0031] S...
Embodiment 2
[0076] Based on the same inventive concept as the deep learning-based method for segmentation and identification of lesions in sugar network images in the foregoing embodiments, such as Figure 4 As shown, the present application provides a deep learning-based sugar network image lesion segmentation and identification system, wherein the system includes:
[0077] An information calling unit 100, the information calling unit is configured to call a pre-stored fundus image data set through an intelligent lesion segmentation identification system, wherein the fundus image data set includes hemorrhage, microvascular tumor, hard exudation, soft exudation, New blood vessel fibroproliferative membrane, abnormal microvascular glucose network lesions;
[0078] a sample expansion unit 200, the sample expansion unit is configured to perform sample expansion of the fundus image data set according to the lesion type to obtain an expanded fundus image data set;
[0079] Input unit 300, the...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More