The Research of Digital Food Experience for Older Population: Innovative Ai-Driven Dietary Intake Reporting and Nutrition Service to Assist Healthy Eating

Older population continues to grow that poses a significant financial burden on healthcare care. Maintaining a healthy eating behavior is a cost effective approach that could ultimately reduce the risk of chronic diseases and reduce healthcare cost. With the rise of artificial intelligence (AI) and machine learning (ML), automating tasks in food and nutrition services is rapidly progressing in the few applicable areas such as integrating food image recognition within mobile technology. Innovative services are made possible, e.g., an app named “享食添糖” developed by that allows users to take a food photo and thus provide a feedback in calorie and nutrition information within ten seconds. However, to reduce errors in the calorie calculations and nutrition information, advanced methods are required in differentiating dish presentation, food preparation methods, and regional changes among a variety of dishes. Older adults are faced with the deterioration of body and mind, and some existing assessment methods for dietary intake often make the elderly feel overwhelm. Self-memory related to dietary assessment methods are challengeable for older adults. Further, due to changes in lifestyle, when being cared by a full-time care facility or relatives, the older adults might not be aware the details upon what and how much of they have eaten. Assisting older adults in reporting dietary intake remain to be unsolved. To complete project goal, this research project is divided into three phases Activities in each phase are planned to be completed in one year. In the first year, we will apply the research methodology of IDEAS, developed by Stanford University Medicine School, to investigate problem-solving approach, to empathize specific seniors’ adults’ physical and mental functions, and to identify key user requirements. In this year, user experience design for older adults will be conducted. We will also include image and doodle recognition and user-centered design. In the second year, we will focus on app development that following software engineering development process, and design user interfaces for automatic image and doodle recognition. After the development of prototype system, the system will be evaluated including expert’s heuristic judgement and pre-test of lead users. In the third year, field usability testing will be conducted under actual user meal conditions to assess the accuracy, efficiency, subjective reaction of three dietary intake reporting approaches embedded in the developed mobile app. A group of seniors (around 105 participants) will be recruited for a randomized comparison, and divided into three equal groups. Assessment was based on breakfast, lunch, and dinner served in an elderly community (e.g., Chang Gung Health and Culture Village). Each of the three approaches featured a unique dietary intake reporting strategy will be evaluated: image-doodle based group (experiment #1 group), doodle-button based (experiment #2 group), and typical user interface with buttons and lists (control group).

Project ID:PB10901-1629
External Project ID:MOST108-2221-E182-008-MY3