Early prediction of acute kidney injury using machine learning algorithms

The application of machine learning algorithms in the medical sector is gaining increased attention in the last few decades. Thus, the main aim of this manuscript is to compare the performance of well-known machine learning (ML) algorithms to a problem in the domain of medical diagnosis and analyze their efficiency in predicting the results. The problem that has been considered in this study is the detection of acute kidney injury (AKI). The ML algorithms are Support Vector Machine (SVM), Neural Network (NN), Deep learning, Decision trees and Naiive Bayes. This research proposed i) an AKI Model: AKI (indicator of renal function) represents a significant risk factor for mortality for patients in ICU, ii) to use analytics to improve clinical decision support by taking advantage of the massive amounts of data and provide right intervention to the right patient at the right time, iii) to use analytics for better care coordination