作者: Christina D. Lee , Junghoon Chae , TusaRebecca E. Schap , Deborah A. Kerr , Edward J. Delp
DOI: 10.1177/193229681200600231
关键词:
摘要: Background: Diet is a critical element of diabetes self-management. An emerging area research the use images for dietary records using mobile telephones with embedded cameras. These tools are being designed to reduce user burden and improve accuracy portion-size estimation through automation. The objectives this study were (1) assess error automatically determined portion weights compared known foods (2) compare between automation human. Methods: Adolescents (n = 15) captured their eating occasions over 24 h period. All beverages served weighed. self-reported sizes one meal. Image analysis was used estimate weights. Data weights, automated portions. Results: For 19 foods, mean ratio weight ranged from 0.89 4.61, 9 within 0.80 1.20. largest lettuce most accurate strawberry jam. children fairly estimates two (sausage links, toast) type aid scrambled eggs) another aid. method jam); however, 95% confidence intervals consistently narrower than human estimates. Conclusions: