Browsing by Author "Nusrat Jahan Khan Shila"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Open Access
    Deep Learning Based Diabetic Retinopathy Detection from Fundus Images with Mobile Application & Web Solution
    (North South University, 2020) Zillur Rahman; Nusrat Jahan Khan Shila; K. M. A. Salam; 1510154642; 1512879642
    Diabetic retinopathy is one of the leading causes of blindness today, and the number of patients is growing at an unprecedented rate as the world’s population increases. At the time of the first diagnosis of diabetes, up to 21% of people with type 2 diabetes have been screened for diabetic retinopathy. As the number of patients grows exponentially, it poses two challenges for the future. The challenges are: access to traditional screening tests will be limited if any modern innovative technology does not bridge the demand-supply gap, and if the traditional system attempts to accommodate a larger number of patients with fewer resources, there is a possibility of error and failing to comply with health guidelines and standards. A deep learning-based, scalable software solution may be a game changer in dealing with those unprecedented challenges. As deep learning achieves state-of-the-art performance in many fields, even outperforming experts, it will ultimately restrict inaccurate and faulty diagnoses, and a robust software solution will democratize access to this health technology. This paper proposes a deep neural network or deep learning-powered automated system for screening and grading diabetic retinopathy with 88% accuracy using 299x299 pixel fundus images using a modular android smartphone application.