Recognizing AR-guided manual tasks through autonomic nervous system correlates: a preliminary study

Optical see-through head-mounted displays (HMD) enable optical superposition of computer-generated virtual data onto the user's natural view of the real environment. This makes them the most suitable candidate to guide manual tasks, as for augmented reality (AR) guided surgery. However, most co...

Full description

Saved in:
Bibliographic Details
Published in2020 IEEE International Symposium on Medical Measurements and Applications (MeMeA) pp. 1 - 6
Main Authors Nardelli, Mimma, Condino, Sara, Ghiasi, Shadi, Callara, Alejandro Luis, Rho, Gianluca, Carbone, Marina, Ferrari, Vincenzo, Scilingo, Enzo Pasquale, Greco, Alberto
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2020
Subjects
Autonomic nervous system
autonomic nervous system (ANS)
Biomedical optical imaging
electrodermal activity (EDA)
Head mounted displays (HMDs)
heart rate variability (HRV)
Manuals
optical seethrough (OST)
Resists
support vector machine (SVM)
Support vector machines
Surgery
vergence-accommodation conflict (VAC)
Visualization
Online AccessGet full text

Cover

More Information
Summary:Optical see-through head-mounted displays (HMD) enable optical superposition of computer-generated virtual data onto the user's natural view of the real environment. This makes them the most suitable candidate to guide manual tasks, as for augmented reality (AR) guided surgery. However, most commercial systems have a single focal plane at around 2-3 m inducing "vergence-accommodation conflict" and "focal rivalry" when used to guide manual tasks. These phenomena can often cause visual fatigue and low performance. In this preliminary study, ten subjects performed a precision manual task in two conditions: with or without using the AR HMD. We demonstrated a significant deterioration of the performance using the AR-guided manual task. Moreover, we investigated the autonomic nervous system response through the analysis of the heart rate variability (HRV) and electrodermal activity (EDA) signals. We developed a pattern recognition system that was able to automatically recognize the two experimental conditions using only EDA and HRV data with an accuracy of 75%. Our learning algorithm highlighted two different physiological patterns combining parasympathetic and sympathetic information.
DOI:10.1109/MeMeA49120.2020.9137309