PIÑATA: Pinpoint insertion of intravenous needles via augmented reality training assistance

Abstract

Conventional needle insertion training relies on medical dummies that simulate surface anatomy and internal structures such as veins or arteries. These dummies offer an interesting space to augment with useful information to assist training practices, namely, internal anatomical structures (subclavian artery and vein, internal jugular vein and carotid artery) along with target point, desired inclination, position and orientation of the needle. However, limited research has been conducted on Optical See-Through Augmented Reality (OST-AR) interfaces for training needle insertion, especially for central venous catheterization (CVC). In this work we introduce PIÑATA, an interactive tool to explore the benefits of OST-AR in CVC training using a dummy of the upper torso and neck; andexplore if PIÑATA complements conventional training practices.. Our design contribution also describes the observation and co-design sessions used to collect user requirements, usability aspects and user preferences. This was followed by a comparative study with 18 participants - attending specialists and medical residents - that performed needle insertion tasks for CVC with PIÑATAand the conventional training system. The performance was objectively measured by task completion time and number of needle insertion errors. A correlation was found between the task completion time in the two training methods, suggesting the concurrent validity of our OST-AR tool. An inherent difference in the task completion time (p =0.040) and in the number of errors (p = 0.036) between novices and experts proved the construct validity of the new tool. The qualitative answers of the participants also suggest its face and content validity, a high acceptability rate and a medium perceived workload. Finally, the result of semi-structured interviews with these 18 participants revealed that 14 of them considered that PIÑATA can complement the conventional training system, especially due to the visibility of the vessels inside the simulator. 13 agreed that OST-AR adoption in these scenarios is likely, particularly during early stages of training. Integration with ultrasound information was highlighted as necessary future work. In sum, the overall results show that the OST-AR tool proposed can complement the conventional training of CVC.