A Novel Mobile App for Personalized Dietary Advice Leveraging Persuasive Technology, Computer Vision, and Cloud...

Area of Science:

• Digital Health
• Nutrition Science
• Human-Computer Interaction

Background:

• Australian Dietary Guidelines (ADG) offer evidence-based nutrition advice, but adherence is low.
• Poor adherence to dietary guidelines increases chronic disease risk.
• Novel technologies are needed to enhance ADG adherence.

Purpose of the Study:

• To describe the development and design of a prototype mobile app for personalized dietary advice based on the ADG.
• To explore the usability of the prototype for real-time, evidence-based self-management of food choices.
• To provide personalized support for adults in Australia aiming to improve dietary habits.

Main Methods:

• Design science paradigm guided iterative development of a progressive web app.
• Integrated Persuasive Systems Design, cognitive behavioral theory, and the ADG.
• Employed a gain-framed approach and image-to-recipe retrieval for dietary assessment; usability assessed via survey and interviews (N=15).

Main Results:

• Prototype features include image-based dietary assessment, gamified food tracking with feedback, goal setting, and ADG-aligned recipes.
• Prototype quality rated "acceptable" (median 3.46/5), with perceived impact on healthy eating rated 3.83/5.
• Gamification and image-based assessment identified as key drivers of positive user experience.

Conclusions:

• A novel, evidence-based prototype mobile app for personalized dietary advice was successfully developed through cross-disciplinary collaboration.
• The detailed development process enhances transparency and offers insights into creating evidence-based health apps.
• This study exemplifies using computer vision for personalized dietary recommendations, with a revised version under development.