ITDS | SPMS

MEDICAL IMAGE DIAGNOSIS USING CONVOLUTIONAL NEURAL NETWORKS FOR BRAIN STROKE AND PNEUMONIA DETECTION

Project Description
Supervisor

This project presents the development and implementation of a web-based medical image diagnosis system using Convolutional Neural Networks (CNNs) for the automated detection of brain stroke and pneumonia. The system was built using an Agile methodology and integrates a Flask-based backend with a responsive frontend developed in HTML, CSS, and JavaScript. A pre-trained VGG16 model was fine-tuned and augmented with custom dense and dropout layers to classify medical images into three categories: brain stroke, pneumonia, and normal. The model was trained on curated datasets of MRI/CT scans for brain stroke and chest X-rays for pneumonia, using an 80/20 train-test split and extensive data augmentation to enhance robustness. The system achieved an accuracy of approximately 90% for pneumonia detection and 85% for brain stroke detection, as evaluated through performance metrics including confusion matrices, ROC curves, and classification reports. The web application allows medical practitioners to upload images via a user-friendly interface and receive real-time diagnostic results. Testing confirmed the system’s reliability and usability, demonstrating its potential as a supportive tool in clinical settings. Future work may include expanding the model to additional diseases, improving integration with electronic health records, and enhancing data privacy measures.

Web App Development: Web Development research in Computer Science and Information Technology focuses on the design, creation, and optimization of websites and web applications that are accessible over the internet or private networks. This research area spans a variety of topics, including front-end and back-end development, web frameworks, web performance, user experience (UX) design, security, and emerging web technologies. Front-end development research explores the design and functionality of the user-facing side of web applications. This involves studying languages such as HTML, CSS, and JavaScript, as well as popular frameworks like React, Angular, and Vue.js, which help create dynamic, responsive, and interactive user interfaces. Researchers in this area often focus on improving the accessibility and usability of websites across different devices and browsers. Back-end development focuses on the server side of web applications, where databases, server logic, and application functionality are managed. Research in this area includes database management, server configuration, APIs (Application Programming Interfaces), and the use of frameworks such as Node.js, Django, and Ruby on Rails. It also involves optimization of server-side processes to handle large-scale traffic, improve load times, and ensure seamless data processing.

The main goal of this research is to create a medical application that utilizes convolutional neural networks to aid in the early detection and diagnosis of brain stroke and pneumonia.

2024/2025

Jan 21, 2026

SASU EMMANUEL OPOKU BEATRICE MOHAMMED NAZIR MARDIYA OWUSU ACHIAW BREMPONG KWAK ADINKRA NKETIAH HERTTY

BSc Information Technology