Abstract
BACKGROUND: Estimation of food portions is a vital skill for dietitians, which is developed during formal nutrition training. Skill development is often accomplished by training with food portion estimation tools. These tools can vary in design but evaluations often reveal them to be limited in their effectiveness and generally impractical for everyday use. The aim of this study was to develop and evaluate an augmented reality (AR) tool for the estimation food portions. METHODS: An online, quasi-experimental, randomised pre-test post-test study was conducted to evaluate the effectiveness of three food portion tools with nutrition students. These tools consisted of an online, AR, and an infographic tool (control). Students tested 10 different food images and were asked to estimate food portion sizes with and without assistance of a portion tool to determine absolute error, relative error, and overall improvement in estimation. RESULTS: A total of 33 participants enrolled in the study with 26 (72.0%) completing the study. The mean absolute error was lowest in the online group (53.0%), followed by AR (59.5%) and control (64.0%). Relative error scores revealed higher accuracy for the AR group (45.5%) followed by online (43.5%), and control group (29.0%). Overall improvement in estimation was highest in the AR group (+12.2%) followed by the online (+11.6%) tool with a decrease seen for the infographic (-1.7%) tool. CONCLUSIONS: The use of technology, notably AR technology, may provide some advantage when training nutrition students in food portion estimation, although further investigation is advised.