Vocal Pain Expression Augmentation to Improve Interaction Accuracy in Virtual Robopatient

Abstract

Palpation is a method use by physicians to phys- ically examine patients using fingers or hands to diagnose any disease or illness. Vocal pain expressions of the patient during palpation are considered as important feedback to assess the conditions. Although recent technological advances has enabled development of medical simulators for physician to train the palpation procedures, incorporating vocal pain expressions to these simulators has been understudied. In this paper, we present a vocal pain expression augmentation for a robopatient to be used in abdominal palpation training. Our virtual robopatient builds upon a virtual abdomen and a face which can render facial pain expressions together with vocal pain expressions. In a user study (N=26), we test the vocal pain augmented virtual robopatient against a system without vocal pain expressions in a palpation task to estimate the maximum pain point within the virtual abdomen. We demonstrate that the vocal pain augmented virtual robopatient leads to statistically significant improvements in localizing the maximum pain without compromising the position estimation time.