CHRONIC KIDNEY DISEASE STAGE CLASSIFICATION USING LABORATORY BLOOD PANEL FEATURES

Abstract

Chronic Kidney Disease (CKD) is a progressive medical condition characterized by
the gradual decline of kidney function, which can lead to severe complications such as
cardiovascular disorders, metabolic imbalances, and end-stage renal failure. Early
detection and accurate staging of CKD are essential for effective treatment and
prevention of disease progression. However, traditional diagnostic methods based on
estimated glomerular filtration rate (eGFR) and clinical assessments may not always
capture subtle variations in kidney function, leading to delayed or imprecise diagnosis.
This project, titled “Chronic Kidney Disease Stage Classification Using Laboratory
Blood Panel Features,” proposes a machine learning-based approach to improve the
accuracy and efficiency of CKD stage classification. The system utilizes laboratory
blood panel data, including features such as serum creatinine, blood urea nitrogen,
hemoglobin levels, sodium, potassium, and other key biomarkers, to analyze patient
health conditions.
Data preprocessing and feature selection techniques are applied to identify the most
relevant attributes influencing CKD progression. Multiple supervised machine
learning algorithms, including Random Forest, Support Vector Machines (SVM),
Gradient Boosting, and Logistic Regression, are implemented to build predictive
models. These models are trained and evaluated using performance metrics such as
accuracy, precision, recall, F1-score, and AUC-ROC to ensure reliable classification.
The experimental results demonstrate that ensemble learning methods, particularly
Random Forest and Gradient Boosting, achieve high accuracy (above 90%) in
classifying CKD stages from Stage 1 to Stage 5. Key features such as serum
creatinine, eGFR, and blood urea nitrogen are identified as significant predictors,
aligning with established medical knowledge