Advancing Chronic Kidney Disease Detection: Improving Biomarkers and CT Imaging for Enhanced Detection

Chronic Kidney Disease (CKD) is one of the biggest health concerns globally, affecting millions. This research aims to enhance the early detection and diagnosis of CKD by utilizing advanced machine-learning models for two different types of data: biomarkers and CT scan images. Each model has been trained and then individually validated. The Random Forest model has given the best average accuracy of 99.4% using different cross-validation methods while the CNN model has reached a validation accuracy of 99.41 %, showing minimum overfitting in each. Thus, both of these models can offer a high level of improvement in accuracy in the detection of CKD if applied separately. Future studies are needed to identify how these models can be incorporated into clinical practice for the betterment of early diagnosis and enhanced patient outcomes.