AUTOMATED DIABETIC RETINOPATHY DETECTION USING CNN ALGORITHM

  • Unique Paper ID: 161696
  • Volume: 10
  • Issue: 5
  • PageNo: 380-383
  • Abstract:
  • Diabetic Retinopathy (DR) is quite common, among a segment of the population who have diabetes. Diabetic retinopathy is a condition that damages the tissues of diabetic individuals and in severe cases, it leads to permanent blindness if not diagnosed properly for a long time. If diagnosed early, consequences can be minimized. In previous work, the detection of DR is done by Machine learning and image processing techniques. In our proposed research work we employ a pre-trained machine learning algorithm based on Convolutional Neural Networks (CNN) to expedite the diagnosis process by analyzing images Based on disease severity, captured images are classified into five categories based on disease severity. As CNN is a fully connected neural network a Deep Learning algorithm, it needs little pre-processing than other classification methods. In our work, we use a recently developed and efficient model named EfficientNet-B5 Efficient net B5 in conjunction with threshold is implemented to increase the performance of our work. The effectiveness of our method in classifying DR cases. Our work has the potential to reduce the likelihood of blindness, for individuals affected by DR.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 10
  • Issue: 5
  • PageNo: 380-383

AUTOMATED DIABETIC RETINOPATHY DETECTION USING CNN ALGORITHM

Related Articles