Data Fusion in the O.R.: Improved Clinical Measurements and Intelligent Patient Monitoring

Data fusion is concerned with combining information from knowledge sources such as sensors to provide a greater understanding of a given situation. Two major classes of fusion techniques in robotics and machine intelligence are (1) those based on probabilistic (such as Bayesian reasoning and the theory of evidence) models and (2) those based on least squares techniques (such as extended Kalman filtering). In other domains, such as medicine, the use of artificial neural networks (ANN) and fuzzy logic have also become important in this context. This contribution aims to give an overview of the data fusion techniques that are prevalent in the operating room (OR). In particular the use of data fusion methodologies to provide (a) improved patient measurements and (b) intelligent patient monitoring are considered.