Clinical decision support system for diagnosis and management of Chronic Renal Failure

• Main Authors: Al-Hyari, Abeer Y., Al-Taee, Ahmad M., Al-Taee, Majid A.
• Format: Conference Proceeding
• Language: English
• Subjects: Accuracy; Algorithm design and analysis; Algorithms; artificial neural networks; chronic renal failure; Classification algorithms; Clinical decision support; data classification; Decision support systems; decision tree; Decision trees; Diagnosis; Diseases; Failure; Kidney; Medical diagnostic imaging; Naïve Bayes; Patients; Prediction algorithms; Source code
• DOI: 10.1109/AEECT.2013.6716440

Chronic Renal Failure (CRF) is a gradual loss of kidney's function over a period of time, ranging from months to years. Unlike other chronic diseases, CRF is not yet thoroughly explored in literature. In this paper, we propose a new clinical decision support system for diagnosing patients with CRF. Several data classification algorithms including Artificial Neural Networks (ANNs), Naïve Bayes and Decision Tree are developed and implemented to diagnose patients with CRF and determine the progression stage of the disease. A clinical dataset of 102 instances is collected from patients' records and used in this study. Performance of the developed CRF diagnosis system is assessed in terms of diagnosis accuracy, sensitivity, and specificity and is evaluated by specialist physicians. Furthermore, the open source Weka software is also used in this study for performance comparison and evaluation purposes. The obtained results showed that the developed decision tree algorithm is the most accurate CRF classifier (92.2%) when compared to all other algorithms/implementations involved in this study.