Deep Learning for Ocular Disease Detection with Android-Integrated Healthcare

Deep Learning for Ocular Disease Detection with Android-Integrated Healthcare

Prof. S. A. Dhanawe    Sushant Khandekar    Tejas Hucche    Prasad Kabade    Omkar Kamreddi

Professor                 B.Tech CSE                 B.Tech CSE       B.Tech CSE           B.Tech CSE

Abstract
The growing incidence of ocular diseases necessitates timely and accurate diagnostic tools to improve patient outcomes. This study introduces a deep learning-integrated Android application designed to recognize ocular diseases from retinal fundus images and connect patients with healthcare providers. Utilizing the VGG-19 convolutional neural network and trained on the ODIR dataset, the system performs high-accuracy binary classification to detect common ocular conditions, effectively addressing class imbalance. The mobile platform leverages Firebase for real-time data synchronization, secure patient records, and seamless doctor-patient communication. Key features include appointment scheduling, multilingual medication reminders, and real-time chat with healthcare professionals. This integrated solution not only enhances diagnostic accuracy but also empowers patients to actively participate in their care, ultimately fostering early detection, timely intervention, and better disease management.

·                  Keywords:

·                  VGG-19

·                  Retinal Fundus Images

·                  ODIR Dataset

·                  TensorFlow Lite (TFLite)

·                  Binary Classification

·                  Firebase

·                  Healthcare Integration

·                  Doctor-Patient Chat

·                  Appointment Scheduling

·                  Real-time Sync

·                  mHealth (Mobile Health)