Home // HEALTHINFO 2018, The Third International Conference on Informatics and Assistive Technologies for Health-Care, Medical Support and Wellbeing // View article

Prediction of Patient Outcomes After Renal Replacement Therapy in Intensive Care

Authors:
Harry Freitas da Cruz
Siegfried Horschig
Christian Nusshag
Matthieu-P. Schapranow

Keywords: Clinical Prediction Model; Renal Replacement Therapy; Machine Learning; Supervised Learning

Abstract:
In order to compensate severe impairments of renal function, artificial, extracorporeal devices have been developed to enable Renal Replacement Therapy. The parameters utilized for this procedure and the specific patient characteristics substantially affect individual patient outcomes and overall disease courses. In this paper, we present a clinical prediction model for outcomes of critically ill patients who underwent a specific form of renal replacement, hemodialysis. For this purpose, we employed two machine-learning models: Bayesian Rule Lists and Multi-Layer Perceptron. To provide more transparency to the perceptron model, we applied mimic learning to its output based on a Bayesian Ridge Regression model. Results show that while the perceptron model outperforms the rule-based classifier, the use of the mimic learning approach enables more thorough model scrutiny by a medical expert, revealing possible model biases, which might have gone unnoticed, a sensitive issue in a high-stakes domain such as medicine.

Pages: 7 to 12

Copyright: Copyright (c) IARIA, 2018

Publication date: October 14, 2018

Published in: conference

ISSN: 2519-8491

ISBN: 978-1-61208-675-0

Location: Nice, France

Dates: from October 14, 2018 to October 18, 2018