Attention Based Multi-Label Classification of Diabetic Retinopathy from Optical Coherence Tomography
D. SEGEVRonen BasriTomer BatashItay ChowersDaniel HarariRivkah LenderJaime LeviYahel ShwartzLiran TiosanoShimon UllmanMeirav Galun
0
Citation
10
Reference
10
Related Paper
Abstract:
Diabetic Retinopathy (DR) is a common complication of diabetes that, in severe cases, can result in blindness. Accurate clinical treatment is imperative to prevent these cases and relies considerably on an exact diagnosis of the various symptoms of DR. We aim to advance DR diagnosis by providing a practical tool to automatically classify Optical Coherence Tomography (OCT) scans for DR and to identify and localize DR-related morphological features within the scans. Our system obtains raw OCT input and only sparse clinical annotations at the volume level, which can be obtained automatically from routine electronic medical records.We developed a novel neural network architecture, OCT-Transformer, that obtains state-of-the-art classification results compared to previous models and does so with limited training data. We base our architecture on an attention mechanism and show this to be the driving factor for the boost in performance. We additionally use our model to locate pixels within the input scans that explain its classification.Fundus (uterus)
Cite
Citations (2)
Cite
Citations (3)
Diabetic retinopathy is an eye disease or condition that can cause vision and blindness in people who have diabetes. It also effects the blood vessels in the retina the sensitivities layer of tissue in the eyes. Diabetic patient needs to get eye check up at least once in year. At early stages diabetic retinopathy may not have any symptoms but finding it in the early stages can protect from the loss of vision. Diabetics ca n be controlled by staying physically active, eating healthy, taking medication regularly.
Cite
Citations (0)
Diabetic retinopathy is a common cause of blindness in the adult population in the United States. Ophthalmologists now have laser and surgical treatment modalities available that can significantly decrease the risk of blindness in the diabetic population. The American Academy of Ophthalmology, through its Diabetes 2000 program, is making a national effort to educate all physicians who care for diabetic patients to recognize the problem and to be aware of the therapies available to prevent blindness.
Modalities
Treatment modality
Cite
Citations (5)
Cite
Citations (10)
Diabetic retinopathy affects 4.2 million people in the United States and is the leading cause of blindness in working-aged people. As the prevalence of diabetes continues to rise, cost-effective interventions to decrease blindness from diabetic retinopathy will be paramount. While HbA1c and duration of disease are known risk factors, they account for only 11% of the risk of developing microvascular complications from the disease. The assessment of environmental risk factors for diabetic eye disease allows for the determination of modifiable population-level challenges that may be addressed to facilitate the end of blindness from diabetes.
Cite
Citations (19)
Cite
Citations (0)
Abstract Diabetic retinopathy has been, and probably remains, one of the four major causes of blindness in the United States [1,2]. Without treatment, eyes that develop proliferative diabetic retinopathy (PDR) have at least a 50% chance of becoming blind within 5 years [3–5]. Appropriate application of treatments that have been developed in the last three decades can reduce this risk of blindness to less than 5% [6]. Medical treatments designed to maximize blood glucose control and reduce the development and progression of retinopathy can further reduce the risk of blindness [7]. This chapter discusses the treatments available, the evidence that the treatments are effective and whether the treatments are widely used.
Cite
Citations (1)
The leading cause for the blindness in the diabetic patients is due to the Diabetic Retinopathy. According to the recent statistics from the US centre for Disease and prevention estimates tat 29.11 million people have the diabetic disease in the United states of America, when it comes to the world population it number goes to the 340 million. The diabetic eye disease will leads to blindness which can be caused by rapid changes in blood vessels. There are many features present in the retina which can act as primary signs of diabetic which can be prevented by this process. This paper Provides you the various techniques and algorithms which helps in diagnosis the effected retina images. This paper also compare the algorithm, reviews, classifiers and techniques of previously proposed in order to develop the efficient algorithms.
Cite
Citations (1)