(P02-047-25) Designing and Testing a Digital National Nutrition Survey System in Taiwan

Objectives: In order to establish data analysis related to diet, nutrition and its correlation with health, the National Nutrition and Health Survey has been conducting for many years in Taiwan. The survey used tools such as a food frequency questionnaire and a 24-hour dietary recall survey to collect information on people's diet intake. This process requires manual and audio recording by interviewers, and much time is needed to ensure the accuracy and completeness of the survey data. Therefore, our goal is to develop a digital survey information system that can assist in the collection of 24-hour dietary recall questionnaire data and verify its effectiveness.

Methods: Based on an information architecture applying artificial intelligence technology for the collection of dietary data, this project aimed to optimize the process of the digital methods for collection of nutritional survey data, improve the logic functions for user interfaces and, expand ingredients and recipes databases, and modify nutritional assessment and data analysis models. Artificial intelligence technology was used in the collection and identification of dietary data of text, voice and food images.

Results: A comprehensive database of recipes, ingredients, and nutrients was established. By using artificial intelligence models, we can accelerate and accurately identify and analyze the nutritional components. According to the expert focus meeting, we modified the process of the current National Nutrition and Health Survey 24-hour dietary recall questionnaire interviewing mode, developed and enhanced the convenience and efficiency functional input logic to improve interviewer-friendly interfaces. A food images/map system was developed to assist the respondents for the input of dietary records.

Conclusions: A prototype web-based digital dietary recall system for nutrition survey has been built and tested. After preliminary field testing, using the new interview mode with food images can help respondents to describe their food intake efficiently, and reduce the time consumption of the survey process.

Funding Sources: This project is supported by Health Promotion Administration, Ministry of Health and Welfare, Taiwan, and National Science and Technology Council, Taiwan.