The graphical presentation of decision support information in an intelligent anaesthesia monitor

This contribution examines the graphical presentation of decision support i nformation generated by an intelligent monitor, named SENTINEL, developed f or use during anaesthesia. Clinicians make diagnoses in real-time during op erations by examining clinically significant trends in multiple signals. SE NTINEL attempts to mimic this decision process by using a system of fuzzy t rend templates. SENTINEL'S implementation of fuzzy trend templates is capab le of providing the dual fuzzy measures of belief and plausibility, which a re derived from the theory of evidence. It is thus capable of generating fa irly rich diagnostic decision support information. However, for SENTINEL to he effective, the visual presentation of this information must be intuitiv e to the anaesthetist, who may not be familiar with the theory of evidence. This paper discusses techniques that are being evaluated to meet the requir ements of the SENTINEL anaesthesia monitor. Specifically, the paper present s methods for highlighting clinically significant trends in physiological ( ur derived) signals by superimposing a coloured band on the signal that ref lects fuzzy output from the intelligent monitor. This paper also discusses the intuitive graphical presentation of binary diagnostic fuzzy measures, i ncluding their further interpretation and presentation as crisp "alarm" and "warning" conditions. (C) 2001 Elsevier Science B.V. All rights reserved.