17 results about "Diabetic retina" patented technology

The invention discloses a diabetes retina image classification method and system based on deep learning, and relates to the technical field of image classification, and the specific steps are as follows: collecting and preprocessing a retina image data set, and constructing a training data set; constructing an initial image classification model based on a deep neural network, and performing iterative training on the initial image classification model by using the training data set to obtain an image classification model; the initial image classification model takes a pre-trained OfficientNet network as a backbone network, and integrates a feature pyramid network structure, a lesion attention module and a classification module; and obtaining a to-be-classified image, preprocessing the to-be-classified image, inputting the preprocessed to-be-classified image into the image classification model, and outputting a classification result. According to the method, the dependence of the model on a specific data source is reduced by using a data enhancement strategy, the diagnostic performance and stability of the model on unseen retina images are improved, and the universality of clinical application is enhanced.

The invention discloses an improved lightweight diabetic retina fundus image small target detection method based on RT-DETR. The detection method comprises the following steps: step 1, constructing a diabetic retina fundus image small target detection model; 2, obtaining a fundus image to be detected, and preprocessing the fundus image; 3, training and optimizing the detection model constructed in the step 1 by using a pre-training set, and inputting the to-be-detected eye fundus image preprocessed in the step 2 into the optimized detection model for detection reasoning; and step 4, outputting an eye fundus image with small target positions, classification and classification confidence. The invention further discloses a detection system for realizing the small target detection method and application of the small target detection method and the detection system, and the application prospect is wide.

The invention discloses an eye fundus image quality control management method for diabetic retinopathy recognition, and the method comprises the following steps: S1, collecting wide-angle eye fundus image data of diabetic retinopathy, and carrying out the preprocessing of a wide-angle eye fundus image; s2, constructing a quality control neural network model based on a ResNet50 network structure; s3, inputting the preprocessed wide-angle eye fundus image into the quality control neural network model, judging whether refractive interstitial opacification exists or not, if yes, returning to output an image quality substandard report, and if not, entering the step S4; s4, on the basis of the quality control neural network model and in combination with a multi-scale feature extraction strategy, capturing a global structure and local details in the wide-angle fundus image, identifying typical features of common disease species of diabetic retina, judging whether the common disease species of the diabetic retina exist or not, if yes, outputting a corresponding disease species prediction report, and if not, outputting a prediction report of the disease species of the diabetic retina. And if yes, judging that the diabetic retinopathy stage is entered.

The present invention relates to a nanoparticle comprising a nanomaterial and at least a first ligand and a second ligand tethered to the nanoparticle. The present invention also relates to a nanoparticle for use as a medicament or diagnostic agent. The present invention also relates to a nanoparticle for use in a method for preventing or treating a disease selected from the group consisting of diabetic nephropathy, glomerulonephritis, glomerular VEGF A dysregulation, endothelial VEGF A dysregulation, diabetic retinopathy, rheumatoid arthritis, age-related macular degeneration and cancer, such as breast cancer. Furthermore, the present invention relates to a method of preparing a nanoparticle.

The invention provides a diabetic retina microangiopathy early diagnosis system based on artificial intelligence, and relates to the technical field of artificial intelligence, and the diabetic retina microangiopathy early diagnosis system specifically comprises an image acquisition module, an image preprocessing module, a feature extraction module, an artificial intelligence diagnosis module, and a result display and report module. Through the trained artificial intelligence algorithm, the extracted features are analyzed, the retina image can be rapidly analyzed, a preliminary diagnosis result is given in a short time, compared with manual diagnosis, the efficiency can be greatly improved, a patient can know the eye condition of himself / herself more rapidly, the detection efficiency is improved, and the detection accuracy is improved. The artificial intelligence system is trained through a large amount of retinopathy image data, various lesion feature modes can be learned, some tiny early lesion signs can be found, misdiagnosis or missed diagnosis caused by human factors is reduced, and the diagnosis accuracy is improved.

The invention provides a new application of a sagittaria sagittifolia extract, a traditional Chinese medicine composition and a preparation method of the sagittaria sagittifolia extract, and the sagittaria sagittifolia extract has a remarkable treatment effect in the aspects of relieving early-stage type 2 diabetic retinopathy and reducing blood sugar and blood fat, and can be used for preparing a medicine for relieving early-stage diabetic retinopathy. The important clinical application value is realized; the traditional Chinese medicine composition is prepared from the following active ingredients in parts by weight: 50 to 70 parts of arrowhead extract, 5 to 15 parts of astragalus extract, 5 to 15 parts of radix pseudostellariae extract, 5 to 15 parts of radix rehmanniae recen extract, 5 to 15 parts of radix scrophulariae extract, 5 to 15 parts of radix salviae miltiorrhizae extract, 5 to 15 parts of radix notoginseng extract, 5 to 15 parts of radix puerariae extract and 5 to 15 parts of semen cassiae extract. The effects of relieving early diabetic retinopathy, reducing blood sugar and blood fat and inhibiting inflammatory microglial cells are obvious; the preparation method is simple, and is suitable for the requirements of large-scale industrial production.

The application provides application of human umbilical cord mesenchymal stem cell-derived Supermeres in preparation of a medicine for treating diabetic retinal fibrosis, and belongs to the technical field of biological medicine; the application research finds that the human umbilical cord mesenchymal stem cell-derived Supermeres can significantly inhibit retinal cell fibrosis, reduce the expression of fibrosis marker proteins alpha-SMA, col-1a1 and Fibronectin, compared with human umbilical cord mesenchymal stem cell-derived EVs, can not only better inhibit the occurrence of fibrosis, but also more significantly inhibit cell apoptosis and promote cell proliferation, so that the technical effect of treating retinal fibrosis is achieved; the human umbilical cord mesenchymal stem cell-derived Supermeres have good application in preparation of the medicine for treating diabetic retinal fibrosis.

The invention provides a diabetes retina vessel segmentation method based on deep learning, and the method comprises the steps: carrying out the processing of a diabetic retinopathy fundus image through a trained optimal UFNet model, and segmenting a fundus vessel image; the UFNet model inherits a Unit + + decoder thought, an SE-CA module is added in each layer of encoder of the Unit + + model, and then the feature is output after passing through a feature fusion module; wherein the SE-CA module is composed of an SE attention module and a CA attention module which are arranged in sequence; the CA attention module generates channel attention features by using two modes of global average pooling and maximum pooling, then performs convolution processing on the two pooling results, and finally obtains a channel weight through a sigmoid function; according to the method, blood vessel segmentation and labeling can be carried out on the diabetic retinopathy image at a higher speed, meanwhile, the accuracy rate is high, the requirement for real-time detection can be met, and the method is suitable for a low-computing-power platform.

The application discloses a diabetic retinopathy feature classification method, system, device and medium, wherein the method comprises: acquiring a diabetic retinal lesion image; inputting the lesion image into a preset channel attention network and a preset spatial attention network respectively to obtain a first feature map output by the channel attention network and a second feature map output by the spatial attention network; fusing the first feature map and the second feature map to obtain a third feature map; inputting the third feature map into a capsule network to obtain an output result of the capsule network; and inputting the output result into a preset classifier to obtain a classification result of the lesion image output by the classifier. The application can improve the accuracy and classification efficiency of diabetic retinal lesion image classification, and better cope with the problems caused by large differences between diabetic retinal lesion images and large amounts of diabetic retinal lesion image data.

The invention provides an application of a GLP-1 (Glucagon Like Peptide-1) similar peptide dimer in preparation of a medicine for preventing and / or treating diabetic complications, and belongs to the technical field of biological medicines. It is found in repairing cell damage of diabetic retinopathy that the GLP-1 similar peptide dimer has the effect of preventing oxidative stress of diabetic retinopathy, the dimer can further remarkably reduce the active oxygen level after the retinopathy is damaged, and the GLP-1 similar peptide dimer has the remarkable effect of preventing and repairing the damage of the diabetic retinopathy. The research result of treating a mouse with diabetic nephropathy by using the GLP-1 similar peptide dimer shows that the GLP-1 similar peptide dimer can effectively treat diabetic nephropathy. Therefore, more choices are provided for clinical medicines for preventing and / or treating diabetic complications.

The invention relates to the technical field of biology, in particular to application of beta-hydroxybutyrylation modification in preparation of a medicine for treating diabetic retinopathy. According to the application, the inducer for promoting beta-hydroxybutyrylation modification is used for preparing the medicine for preventing and / or treating diabetic retinopathy. The application is based on an in-vitro experiment, and proves that after beta-hydroxybutyrate intervention, beta-hydroxybutyrate is modified by beta-hydroxybutyrylation, so that apoptosis of a mouse retinal ganglion cell strain induced by an advanced glycosylation end product can be reduced; western blot proves that beta-hydroxy butyrylation modification of cells in a ganglion cell layer and a photoreceptor cell layer is induced in the retina in a diabetic animal body after beta-hydroxybutyrate intervention, which shows that the diabetic retina can be prevented and treated by inducing and promoting beta-hydroxy butyrylation modification; the blank between the beta-hydroxyl butyrylation modification and the diabetic retinopathy in the existing research is filled.

The application discloses a kind of small sample-oriented diabetic retinopathy classification method and system, belong to image processing and medical field, wherein, classification method is by pre-training stage based on lightweight RexNet channel optimization, fundus image data preprocessing stage, data enhancement stage, model fine-tuning training stage based on transfer learning, system implementation stage composition;The application solves the problem of medical image data sample and patient privacy, breaks the barrier of insufficient DR diagnostic ability in basic community.At the same time, combined with transfer learning, pre-training is carried out in ImageNet to obtain a pre-trained model.The data is enhanced to improve the accuracy of the diabetic retinopathy classification and detection system, reduce the computing cost, assist the basic community in clinical diagnosis of diabetic retinopathy, and conduct prevention and screening diagnosis of diabetic retinopathy.

The invention discloses a focus segmentation method for a diabetic retinopathy fundus image. The focus segmentation method comprises the following steps: S1, collecting wide-angle fundus image data of diabetic retinopathy; and S2, based on a multi-scale feature extraction strategy, capturing a global structure and local details in the wide-angle fundus image, identifying typical features of common disease species of diabetic retina, judging whether the common disease species of fundus exist or not, if so, outputting a corresponding disease species prediction report, and if not, performing diabetic retinopathy staging on the wide-angle fundus image to obtain a diabetic retinopathy staging result. Outputting the staging category with the highest probability; s3, constructing and training a focus segmentation model and an optic cup and optic disc segmentation model; s4, segmenting the focus of the wide-angle fundus image through the focus segmentation model and the optic cup and optic disc segmentation model, and performing pixel-level labeling on the optic cup and optic disc of the focus; and S5, generating a corresponding screening evaluation report according to the staging category in the step S2 and the focus segmentation result in the step S4.

The invention discloses application of the M2 type microglial cell exosome. The invention discloses an application of an M2 type microglial cell exosome in preparation of a product for treating diabetic retina neuropathy. In combination with a specific embodiment, the M2 type microglial cell exosome (M2-Exos) is separated and proved to promote the survival and axon growth of primary RGCs in vitro. In an STZ-induced diabetic mouse model, in-vivo injection of M2-Exos prevents RGC loss by inhibiting a pro-inflammatory mediator and enhancing an anti-inflammatory factor, so that ganglion cell complex thickness and visual ability are retained. The axon regeneration promoting ability of the M2-Exos is further verified in an optic nerve crushing animal model. The discovery shows that the M2-Exos is a potential neuroprotection and axonal regeneration promotion tool for treating the diabetic retina neuropathy, and the M2-Exos can be used for preparing products for treating the diabetic retina neuropathy.

The invention discloses a diabetic retinopathy fundus image stage prediction method, which comprises the following steps of S1, acquiring diabetic retina image data, and preprocessing a diabetic retina image; s2, inputting the pre-processed diabetic retina image into a pre-established quality control neural network model, judging whether refractive interstitial opacification exists or not, if yes, returning and outputting an image quality substandard report, and if no refractive interstitial opacification exists, combining a multi-scale feature extraction strategy to extract the refractive interstitial opacification of the diabetic retina image; capturing a global structure and local details in the diabetic retina image, identifying typical features of common disease species of the diabetic retina, judging whether common eye fundus disease species exist or not, if yes, outputting a corresponding disease species prediction report, and if not, entering diabetic retinopathy staging prediction; and S3, performing retinopathy staging on the wide-angle fundus image through a pre-trained staging neural network model, and outputting a staging category with the highest probability.

This application provides a method and system for segmenting hard exudates in diabetic retinopathy, relating to the field of image segmentation technology. The method involves determining the image complexity of a color fundus image; segmenting the color fundus image into multiple exudation regions of hard exudates based on the distribution characteristics of hard exudates in the diabetic retina; determining the edge contrast of each exudation region; determining the variability index of diabetic retinopathy based on all edge contrasts and image complexity; determining the convolution fusion boundary of all exudation regions based on the variability index and the boundary characteristics of each exudation region; and performing convolution fusion segmentation on all exudation regions of hard exudates based on the convolution fusion boundary to obtain multiple segmented regions of hard exudates in the color fundus image. This application can achieve convolution fusion segmentation of hard exudates in diabetic retinopathy, thereby improving the segmentation accuracy of hard exudates in diabetic retinopathy.

The invention discloses a method, a system and equipment for evaluating the treatment effect of drugs on diabetic retina fibrosis diseases. The inventor finds that the EGFR kinase inhibitor or the ROCK inhibitor plays a role in treating the diabetic retina fibrosis disease by regulating the expression of YAP or CCN2 for the first time; the YAP inhibitor plays a role in treating diabetic retina fibrosis diseases by regulating expression of CCN2. On the basis, the invention provides the method, the system and the equipment for evaluating the treatment effect of the medicine on the diabetic retina fibrosis disease, so that a doctor is assisted in evaluating the treatment effect of a patient suffering from the diabetic retina fibrosis disease, and a clinician is guided to formulate an individualized treatment scheme for the patient.