Background Diet ID™ is a digital, image-based dietary assessment tool that uses pattern recognition and comparisons of meal images to estimate overall dietary intake and diet quality. Objective To examine the relative validity of Diet ID™ in measuring Healthy Eating Index-2020 (HEI-2020) total and component scores. Methods In this cross-sectional study, participants completed Diet ID™ and the Diet History Questionnaire II, a semiquantitative food frequency questionnaire. Spearman's rho correlations examined relative validity. Wilcoxon signed-rank tests, cross-classification, and Bland–Altman plots assessed agreement. Results The HEI-2020 total scores were moderately correlated ( r s = 0.46, p < 0.001) and most components were weakly correlated ( r s = 0.12–0.34, p < 0.05). There was poor agreement between the dietary assessment methods, as evidenced by significant differences in most scores ( p < 0.001), fewer than half of the participants being classified into the same tertile, and Bland–Altman plots. Conclusion Further work is needed to improve Diet ID™ to measure HEI-2020 total and component scores.
Stott et al. (Tue,) studied this question.