Abstract: Chronic Kidney Diseases are undoubtedly one of many fatal diseases that are very difficult to diagnose. With early diagnosis and treatment, it is possible to slow or stop the progression of chronic kidney disease (CDK). However, the most basic kidney screening tests available, such as the Blood Urea Nitrogen (BUN)/creatinine ratio test, cost $25 or more. This disease remains a leading cause of end-stage kidney disease thus requiring renal replacement therapy with dialysis or kidney transplantation. In this work we developed a model to predict the various stages of kidney diseases using deep belief neural network using clinical data. The model was trained using dataset containing 400 patient records. The attributes used for building our model includes; age, blood pressure, albumin, red blood cells, pus cells, pus cells clumps, hemoglobin, White blood cell count, Red blood cell count, etc. Object-oriented design methodology was used for system development. Localization model was used to identify quality kidney images. Deep belief neural network was used to analyze the image of the kidney and predict the status of kidney health. System was implemented in python programming language. The system successfully classified kidney disease dataset into CKD and non-CKD with 98% overall accuracy when the model was tested with a set of data that were not used during the training process.
Keywords: Deep neural networks, chronic kidney disease, supervised machine learning.
Works Cited:
Aloy-Okwelle C, O.T. Olise , O.P Nweke " A Deep Belief Neural Network Model for Predicting the Early Phases of Chronic Kidney Disease ", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 12, no. 10, pp. 1-10, 2023. Crossref https://doi.org/10.17148/IJARCCE.2023.121001
|
DOI:
10.17148/IJARCCE.2023.121001
