MEDX-Vision Smart Diagnosis Through Deep Learning

Abstract

Medx-Vision is an AI-powered mobile application that simplifies chest disease detection by analyzing X-ray images and providing easy-to-understand diagnostic results. Using a Convolutional Neural Network (CNN) trained on the NIH Chest X-ray dataset, the system identifies conditions like pneumonia and cardiomegaly with high accuracy. The backend, built with Flask, preprocesses images and returns predictions with confidence scores, which are formatted into laymanfriendly messages. The Android app, developed using Jetpack Compose, enables users to upload or capture images and view results through a clean, intuitive interface. Designed for accessibility, Medx-Vision bridges the gap between complex medical AI and everyday users, making early diagnosis more available in underserved areas.