부산대학교 도서관

Caregivers and nurses have, additionally to the task of giving care, the responsibility to dispense and distribute medication to patients according to a medication plan. Due to the growing shortage of skilled nursing staff, the burden on individual nurses is steadily increasing. This increased cognitive and physical load also leads to an increased potential for error, especially in the aforementioned task of sorting pills. Given the potentially fatal outcome of errors in medication, this work presents a novel approach supporting the pill sorting task using augmented reality (AR). By leveraging artificial intelligence (AI), the current progress of the preparation task is monitored, and potential errors are reported instantly. For evaluation purposes, laymen and trained nurses had been invited to evaluate and rate the system. Participants first used a printed medication plan to execute the pill sorting task, followed by the AR-application. Simultaneously, distractions were simulated by applying cognitive overload via dual-tasking. In effect, a significantly lower error rate was observed when using the AR system when compared to conventional methods. A system usability score above average also underpins the feasibility of this concept on a participantive scale.