Patient Infusion Pattern based Access Control Schemes for Wireless Insulin Pump System

摘要: Wireless insulin pumps have been widely deployed in hospitals and home healthcare systems. Most of them limited security mechanisms embedded to protect from malicious attacks. In this paper, two attacks against pump systems via wireless links are investigated: a single acute overdose with significant amount medication chronic small extra over long time period. They can be launched unobtrusively may jeopardize patients’ lives. It is very urgent patients these We propose novel personalized patient infusion pattern based access control scheme (PIPAC) for pumps. This employs supervised learning approaches learn normal patterns terms the dosage amount, rate, infusion, which automatically recorded logs. The generated regression models used dynamically configure safe range abnormal identification. model includes sub bolus (one type insulin) detection basal rate detection. proposed algorithms evaluated real pump. evaluation results demonstrate that our able detect high success rate.