A Hybrid ArtInt Model to Predict Diabetic Retinopathy

Abstract

Diabetic retinopathy (DR) is retinal blood vessel damage caused by diabetes (DED). Untreated, the condition can cause blindness. Early visual impairment can be averted by monitoring and treating diabetes. Many scientists have developed machine learning techniques to detect DR earlier. Model results proved their usefulness. A WHO research predicts that 40–45% of the world's 347 million diabetics have DR. Early DR diagnosis and therapy can substantially slow visual loss. Current DR detection approaches entail skilled doctors manually analysing digital fundus images, which can cause miscommunication and postpone therapy. This study grades DR intensity as mild, moderate, or severe. A new AI model uses support vector machines and Vgg-16 to classify DR. The results suggest that integrating both models yields better outcomes than standard models.