Medico: AI Model to Analyze X-ray Images

Abstract

The need for faster and more accurate medical image interpretation has driven the inclusion of Artificial Intelligence (AI) in radiology workflows. The paper offers an AI-assisted diagnostic platform for the evaluation of chest X-ray and MRI images using a fine-tuned DenseNet121 deep convolutional neural network. The model labels medical images into two classes: Normal and abnormal, and is implemented through an easy-to-use web interface using Streamlit. The system has secure user authentication, real-time image upload, and interactive visualization of results, which make it applicable for clinical and educational purposes. It also emulates lung condition parameters from classification confidence to add extra interpretative information. The application is made with accessibility and ease of use in mind, and it shows the potential of AI in supporting radiologists and healthcare professionals by providing preliminary diagnostic assistance, particularly in resource-constrained environments. Future development involves expanding to multiclass diagnosis and incorporating real-world clinical feedback to make it more accurate and usable.