AI-BASED MEDICAL IMAGE PREDICTOR USING CONVOLUTIONAL NEURAL NETWORKS FOR MULTI-DISEASE DETECTION

Abstract

AI is developing very fast and has really changed medical imaging, giving us new ways to be more correct and quicker at diagnosing illnesses. Looking at medical pictures like X-rays, CT scans and MRIs generally needs a specialist with lots of training, and it can take a while, which could mean a diagnosis is delayed. This is a big issue in many places, in particular if finding a skilled radiologist is hard. We’ve created an AI system to help doctors and other healthcare staff examine medical images. It uses a sophisticated form of learning called deep learning, and in particular, Convolutional Neural Networks (CNNs), to spot things like broken bones, pneumonia and brain tumours. It's a program you use on the internet: you upload a picture and it gives you its idea of what’s wrong, and how sure it is. When we tried it out, it did a good job with all sorts of medical images, though how accurate it is affected by the image’s quality and how much data it has been trained on. But even with these difficulties, it looks like a good way to help with diagnosis, not to replace doctors. In short, this shows how AI can be included in the usual way medical images are used to lessen the amount of work done by medical staff, work more quickly and give doctors more help when they are deciding on treatment.