MedDietAgent is a context-aware mobile computing system aiming at harmonizing one individual’s dietary choises with national dietary guidelines, and more specifically, dietary patterns consistent with the Greek culinary culture, emphasizing the Mediterranean diet. The main innovative element of the MedDietAgent consists in the employment of reinforcement learning techniques, which provide dynamic recommendations by monitoring the user’s environment and targeting at optimizing a reward function.
Thus, MediDietAgent acts as an agent which progressively aligns the daily dietary choices to the healthy Mediterranean style, by continuously monitoring the user’s dietary history along with his/her pertinent context and according to the extent at which the personal nutritional goals have been achieved. The dietary state of an individual, which formulates the reinforcement leaning algorithm’s input, is captured by combining multimodal food recording and automated food nutritional analysis.
In particular, MedDietAgent employs machine learning for natural language processing and compute vision in order to enhance the granularity level of food recording and analysis, minimizing, in parallel, the time complexity of the food recording procedure.
Accordingly, the MedDietAgent data representation model will be based on holographic imaging technologies, aiming at increasing the commitment to the proposed application, and eventually. the degree of compliance with the Mediterranean dietary pattern. MedDietAgent’s functionality will be tailored to the user’s personal preferences and those environmental factors affecting his / her choices and decisions (at home, in the work environment, in the restaurant, on holidays).
MedDietAgent targets healthy adults or chronic disease patients taking into account additional medical parameters (e.g. blood pressure, blood glucose).
Project Τitle: An Intelligent Context-aware System for Harmonizing Individual Dietary Choices with Mediterranenan Diet in Greece based on Reinforcement Learning
Project Duration: 29-7-2021 to 28-11-2023
Project URL: http://www.meddietagent.gr/
Project Framework & Funding: Operational Programme Competitiveness, Entrepreneurship and Innovation 2014-2020 (EPAnEK)
Scientific coordinator (UNIWA): Maria C. Giannakourou
Consortium – Partners: University of Ioannina (Co ordinator)/Apart/DBC
Research Publications / Results / Patents
|
In Peer reviewed journals |
|
|
1 |
Konstantakopoulos, F.S., Georga, E.I. & Fotiadis, D.I., A novel approach to estimate the weight of food items based on features extracted from an image using boosting algorithms. Sci Rep 13, 21040 (2023). https://doi.org/10.1038/s41598-023-47885-0. |
|
2 |
Mavrokotas, K. I., Georga, E. I., Kanellou, A., Papaloukas, C., & Fotiadis, D. I., An Intelligent Dietary Guidance System for Promoting Mediterranean Diet Adherence based on Deep Reinforcement Learning, to be submitted in a scientific journal. |
|
In scientific conferences |
|
|
1 |
Mavrokotas, K. I., Georga, E. I., Papaloukas, C., & Fotiadis, D. I., A Nutrition Recommendation System based on Reinforcement Learning, 46th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2024 (submitted on Feb 2nd, 2024). |
|
2 |
Konstantakopoulos, F. S., Georga, E. I., Revelou, P., Giannakourou M. & Fotiadis, D. I., A Deep Ensembled Model for Multiclass Semantic Segmentation of Food Images, 46th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2024 (submitted on Feb 2nd, 2024). |
|
3 |
Konstantakopoulos, F. S., Sfakianos, M., Georga, E. I., Mavrokotas, K., Chalatsis, K., Zapadiotis, C., Panousi, A., Plimakis, S., Eleftheriou, S., Kanellou, A. & Fotiadis, D. I., MedDietAgent: An AI-based Mobile App for Harmonizing Individuals’ Dietary Choices with the Mediterranean Diet Pattern, 46th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2024 (submitted on Feb 2nd, 2024). |
|
4 |
Konstantakopoulos, F. S., Georga, E. I., Tachos, N. S., & Fotiadis, D. I., ‘Weight Estimation of Mediterranean Food Images using Random Forest Regression Algorithm’, 45th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2023 |
|
5 |
Konstantakopoulos, F. S, Georga, E. I., Eleftheriou, S., Revelou, P., Kanellou, A., Giannakourou, M., & Fotiadis, D. I., An automated nutritional assessment system based on food images. In ECO 2023 – 30th European Congress on Obesity, May 17-20, 2023. Obesity facts. |
|
6 |
Kanellou, A., Eleftheriou, S., Vlassopoulos, A., Kapsokefalou, M., Konstantakopoulos, F., Giannakourou, M., Houhoula, D., Georga, E., Fotiadis, D., Halvatsiotis, P., & Revelou, P., Developing an updated Greek food composition database for a dietary mobile app – The MedDietAgent project pilot. In ECO 2023 – 30th European Congress on Obesity, May 17-20, 2023. Obesity facts. |
|
7 |
Konstantakopoulos, F. S., Georga, E. I., & Fotiadis, D. I. (2022). Multiclass Semantic Segmentation of Mediterranean Food Images. In EAI PervasiveHealth 2022 – 16th EAI International Conference on Pervasive Computing Technologies for Healthcare (pp. 1-11). Springer. |
