Development of User-Drawn Doodles for Communication and Reporting of Dietary Intake in Health Management

Abstract

The use of machine learning to support health-related activities is yet to be adequately explored. We present a concept of using user-drawn doodles for communication and reporting of food item and its portion size. This paper aims to devise innovative applications for existing ML-based services. A prototype mobile app is developed and demonstrated. We invite three senior dietitians to evaluate its potential. Analysis is proceeded in used of affinity diagramming. The expert evaluation concluded the proposed system has good potential for promoting food class and macro-nutrient education among children. This research presents an innovative use of existing AI services helping to communicate food intake related information in nutrition education. Our future work will conduct user evaluation, integrate expert and user evaluations to improve app experience and usage, and enhance system functionality.