Ai-powered web-based pressure ulcer staging system

This thesis aims to develop a web-based application for the healthcare sector using deep learning-based image classification methods to assist in the identification of pressure ulcers. In this context, a Vue.js-based web application has been designed, enabling users to accurately identify different stages of pressure ulcers and view the results of AI-supported classification. Additionally, the performance of various deep learning models (InceptionV3, ResNet50, DenseNet121, etc.) for image classification has been investigated. These models have been evaluated based on metrics such as accuracy, precision, recall, and F1 score. The models were assessed based on their ability to classify pressure ulcers into four distinct stages and make correct predictions for each stage. By integrating the deep learning model with technologies like Vue.js and TensorFlow.js into a web environment, the application enables users to upload pressure ulcer images in real-time and receive accurate classifications supported by artificial intelligence. The web application is designed with a user-friendly interface and mobile-responsive features, aiming to assist healthcare professionals in the patient evaluation process. Built with the Vue.js framework, the application adheres to responsive design principles, providing practical solutions for both patients and healthcare providers. It includes essential features like image upload, visualization of classification results, and storage of historical data. Furthermore, usability testing based on user feedback has been conducted, confirming that the application operates securely and effectively with high-accuracy models. Recommendations have been proposed for ensuring the safe and transparent use of AI models in clinical environments. The system is expected to support healthcare professionals in decision-making, thereby contributing to the improvement of clinical processes and enabling faster and more accurate patient evaluations. In conclusion, the combination of deep learning-based models in healthcare and web application development using Vue.js represents a significant advancement in digital health technologies, facilitating the faster implementation of technological innovations in hospital settings.