Purpose Effective training of extracorporeal membrane oxygenation (ECMO) cannulation is key to fighting the persistently high mortality rate of ECMO interventions. Though augmented reality (AR) is a promising technology for improving information display, only a small percentage of AR projects have addressed training procedures. The present study investigates the potential benefits of AR-based, contextual instructions for ECMO cannulation training as compared to instructions used during conventional training at a university hospital. Methodology An AR step-by-step guide was developed for the Microsoft HoloLens 2 that combines text, images, and videos from the conventional training program with simple 3D models. A study was conducted with 21 medical students performing two surgical procedures on a simulator. Participants were divided into two groups, with one group using the conventional instructions for the first procedure and AR instructions for the second and the other group using instructions in reverse order. Training times, a detailed error protocol, and a standardized user experience questionnaire (UEQ) were evaluated. Results AR-based execution was associated with slightly higher training times and with significantly fewer errors for the more complex second procedure ($$p<0.05$$ p < 0.05 , Mann–Whitney U). These differences in errors were most present for knowledge-related errors, resulting in a 66% reduction in the number of errors. AR instructions also led to significantly better ratings on 5 out of the 6 scales used in the UEQ, pointing to higher perceived clarify of information, information acquisition speed, and stimulation. Conclusion The results extend previous research on AR instructions to ECMO cannulation training, indicating its high potential to improve training outcomes as a result of better information acquisition by participants during task execution. Future work should investigate how better performance in a single training session relates to better performance in the long run.

For an in-depth, AOI-based analysis of mobile eye tracking data, a preceding gaze assignment step is inevitable. Current solutions such as manual gaze mapping or marker-based approaches are tedious and not suitable for applications manipulating tangible objects. This makes mobile eye tracking studies with several hours of recording difficult to analyse quantitatively. We introduce a new machine learning-based algorithm, the computational Gaze-Object Mapping (cGOM), that automatically maps gaze data onto respective AOIs. cGOM extends state-of-the-art object detection and segmentation by mask R-CNN with a gaze mapping feature. The new algorithm’s performance is validated against a manual fixation-by-fixation mapping, which is considered as ground truth, in terms of true positive rate (TPR), true negative rate (TNR) and efficiency. Using only 72 training images with 264 labelled object representations, cGOM is able to reach a TPR of approx. 80% and a TNR of 85% compared to the manual mapping. The break-even point is reached at 2 hours of eye tracking recording for the total procedure, respectively 1 hour considering human working time only. Together with a real-time capability of the mapping process after completed training, even hours of eye tracking recording can be evaluated efficiently. (Code and video examples have been made available at: https://gitlab.ethz.ch/pdz/cgom.git)

No abstract

No abstract