Search

The identification of subsurface structures during resection wound repair is a challenge during minimally invasive partial nephrectomy. Specifically, major blood vessels and branches of the urinary collecting system need to be localized under time pressure as target or risk structures during suture placement. This work presents concepts for AR visualization and auditory guidance based on tool position that support this task. We evaluated the concepts in a laboratory user study with a simplified, simulated task: The localization of subsurface target points in a healthy kidney phantom. We evaluated the task time, localization accuracy, and perceived workload for our concepts and a control condition without navigation support. The AR visualization improved the accuracy and perceived workload over the control condition. We observed similar, non-significant trends for the auditory display. Further, clinically realistic evaluation is pending. Our initial results indicate the potential benefits of our concepts in supporting laparoscopic resection wound repair.

Complex bi-manual tasks often benefit from supporting visual information and guidance. Controlling the system that provides this information is a secondary task that forces the user to perform concurrent multitasking, which in turn may affect the main task performance. Interactions based on natural behavior are a promising solution to this challenge. We investigated the performance of these interactions in a hands-free image manipulation task during a primary manual task with an upright stance. Essential tasks were extracted from the example of clinical workflow and turned into an abstract simulation to gain general insights into how different interaction techniques impact the user's performance and workload. The interaction techniques we compared were full-body movements, facial expression, gesture and speech input. We found that leaning as an interaction technique facilitates significantly faster image manipulation at lower subjective workloads than facial expression. Our results pave the way towards efficient, natural, hands-free interaction in a challenging multitasking environment.

The paper focuses on a validation study for a macro-cognitive model of mental workload of airport traffic controllers (ATCOs). This model constitutes a means for quantifying and analysing the distribution of ATCOs workload over time dependent on different traffic scenarios. Workload is modelled via the amount of chunks in working memory. In a one-factor experimental Simulator-Study workload ratings were gathered with different traffic load, using a modified RSME scale within and NASA-TLX after each scenario. The data are analysed and discussed concerning the successful experimental manipulation and the estimation of goodness of fit of the model and experimental data. Therefore the average workload ratings of the participant and the model within each scenario as well as the distribution of workload over time for each participant-model pair are compared. The developed model can serve as a tool for the design of adaptive automation and supporting systems and can help to better understand workload dynamics of airport traffic controllers.