Currently, research on the artificial pancreas (AP) has increased exponentially. This is due to a number of factors including the significant advancement of both the hardware and algorithms associated with the AP. Several AP prototypes have been validated in controlled inpatient studies and a number of trials in home setting are currently being conducted. The domiciliary use of the AP requires, besides an efficient controller, the ability to face the daily life conditions with the safety mechanisms that will allow it to be used without additional risk for the patient. Objective: This study aims to provide the basic infrastructure required by the AP regarding safety. This involves the classification of the various safety issues encountered by AP and a hierarchical system for resolving these issues. Methodology: All safety issues regarding the AP were assessed inclusive of anomalies that may occur due to changing physiological conditions within the patient, as well as, hardware malfunctions, maintenance, and inherent error. Issues were then categorized based on time span, the ability to be detected with available sensors and, more challenging situations that require additional information to be detected. Conclusion: The taxonomy provided will allow subsequent investigations to be done in a structured format. Dealing with the most impending safety issues and then following with less risky situations. It should be noted that all these safety mechanisms should be functional in an AP suitable for unsupervised use
DUGiDocs