A Technology-Driven Approach to Dietary Monitoring: Validation Support, and Effectiveness Assessment for Nutrient Estimation via AI-powered Food Image Analysis

Poor dietary habits and limited nutrition awareness are major contributors to long-term health risks. Scalable, personalized digital tools are needed to promote healthier eating behaviors, and AI–based applications offer a practical solution.

This randomized, controlled interventional study will evaluate the short-term feasibility and effectiveness of an AI-powered mobile application for personalized nutrition tracking among students of Manipal Academy of Higher Education (MAHE), Manipal. The nutrition tracker, developed by Dure Technologies in collaboration with MAHE, is supported by a comprehensive, quantified database of mess meals across the MAHE campus to enable accurate nutrient analysis of meals captured by participants.

Participants will be randomized into two groups for a 3-month period:

• Experimental group: Dietary intake recorded through the AI-powered mobile application.

• Control group: Dietary intake recorded using a Google Form with the same question format. A nutrition pamphlet will be provided for the study participants.

Primary outcomes include the feasibility of routine dietary tracking and personalized nutrition feedback using the AI-based app, as well as the accuracy of the AI model in estimating nutrient intake and identifying dietary imbalances. Findings will generate preliminary data to inform future scale-up and real-world implementation of the application.