ITDS | SPMS

Machine Learning-Based Prediction of Chronic Kidney Disease (CKD)

Project Description
Supervisor

Chronic kidney disease (CKD) places a significant strain on the healthcare system due to its growing patient population, high likelihood of progressing to end-stage renal disease, and associated risks of illness and death. This study aims to develop a machine learning model that leverages comorbidity and medication data from Kaggle Database to predict the onset of CKD within 6 or 12 months before diagnosis, thereby estimating its prevalence in the population. A total of 18,000 individuals diagnosed with CKD and 72,000 without a CKD diagnosis were selected through propensity score matching. Data on demographics, medications, and comorbidities collected over a two-year observation period were used to train the predictive models. Among the various methods tested, the Convolutional Neural Network (CNN) model achieved the highest performance, with GHROC scores of 0.957 and 0.954 for predicting CKD 6 months and 12 months in advance, respectively. Key predictors identified by tree-based models included diabetes mellitus, age, gout, and medications such as sulfonamides and angiotensin-related drugs. The proposed model may serve as a valuable tool for healthcare policymakers by forecasting CKD trends, enabling early identification and monitoring of at-risk individuals, improving resource allocation, and supporting more personalized patient car.

Mobile App Development: Mobile App Development research in Computer Science and Information Technology focuses on the design, development, and optimization of software applications for mobile devices such as smartphones and tablets. This area covers a wide range of topics, including user interface design, cross-platform development, mobile operating systems, performance optimization, and security. One key aspect of research in this area is the development of efficient algorithms and frameworks that enable seamless cross-platform compatibility. Researchers explore ways to create apps that function smoothly on multiple operating systems, such as Android and iOS, using a single codebase, reducing development time and costs. This is often achieved through the use of frameworks like React Native, Flutter, and Xamarin. User experience (UX) and user interface (UI) design are critical components of mobile app development research. Scholars in this field investigate how to create intuitive, responsive, and engaging interfaces that improve usability and enhance the overall user experience. This includes studying interaction patterns, accessibility, and how users interact with mobile apps across different devices and environments.

To develop predictive machine learning models capable of forecasting the onset of Chronic Kidney Disease (CKD), delivered through a Flutter-based desktop application for use by healthcare professionals.

2024/2025

Feb 11, 2026

1. SAMUEL KWESI MENSAH (UEB3257423) 2. AKAYIRI BERNARD (UEB3278423) 3. NUAKO ABIGAIL (UEB3218421) 4. MUSAH DAKIYA (UEB3238923)

BSc Information Technology