Artificial Intelligence in Nephrology: Core Concepts, Clinical Applications, and Perspectives

Am J Kidney Dis. 2019 Dec;74(6):803-810. doi: 10.1053/j.ajkd.2019.05.020. Epub 2019 Aug 23.

Abstract

Artificial intelligence is playing an increasingly important role in many fields of medicine, assisting physicians in most steps of patient management. In nephrology, artificial intelligence can already be used to improve clinical care, hemodialysis prescriptions, and follow-up of transplant recipients. However, many nephrologists are still unfamiliar with the basic principles of medical artificial intelligence. This review seeks to provide an overview of medical artificial intelligence relevant to the practicing nephrologist, in all fields of nephrology. We define the core concepts of artificial intelligence and machine learning and cover the basics of the functioning of neural networks and deep learning. We also discuss the most recent clinical applications of artificial intelligence in nephrology and medicine; as an example, we describe how artificial intelligence can predict the occurrence of progressive immunoglobulin A nephropathy. Finally, we consider the future of artificial intelligence in clinical nephrology and its impact on medical practice, and conclude with a discussion of the ethical issues that the use of artificial intelligence raises in terms of clinical decision making, physician-patient relationship, patient privacy, and data collection.

Keywords: Deep learning; IgA nephropathy (IgAN); algorithms; artificial intelligence; artificial neural network (ANN); augmented intelligence; big data; convolutional neural networks (CNN); hemodialysis; kidney disease; machine learning; medical artificial intelligence (MAI); neural network; review; transplantation.

Publication types

  • Review

MeSH terms

  • Algorithms*
  • Artificial Intelligence*
  • Female
  • Forecasting
  • Humans
  • Machine Learning
  • Male
  • Nephrology / methods*
  • Nephrology / trends
  • Quality Improvement*