Somaye Norouzi1, Azade Kamel Ghalibaf2, Samane Sistani1, Vahideh Banazadeh3, Fateme Keykhaei3, Parisa Zareishargh3, Fateme Amiri3, Mohsen Nematy4, Kobra Etminani1. 1. Department of Medical Informatics, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran. 2. Department of Medical Informatics, Students Research Committee, Mashhad University of Medical Sciences, Mashhad, Iran. 3. Department of Nutritional Sciences, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran. 4. Associate Professor in Clinical Nutrition, Chairman: Journal of Fasting and Health, Head: Department of Nutrition, School of Medicine, Metabolic Syndrome Research Center, Mashhad University of Medical Sciences, Mashhad, Iran.
Abstract
BACKGROUND: As a prevalent metabolic disease, diabetes has different side effects and causes a wide range of co morbidity with a high rate of mortality. There is a need for certain interventions to manage this disease. Iranians usually have three main meals a day. Considering the special needs of diabetic patients and the possibility of hypoglycemia between the main meals, it is essential for these patients to eat something as a snack. Considering these conditions and the society's orientation towards modern technologies such as smart phones, designing mobile-based nutrition recommender systems can be helpful. METHODS: The snack recommender system is a knowledge-based smart phone application. This study has focused on the development of a recommender system that combines artificial intelligence techniques and makes up a knowledge base according to the guidelines posed by the American Diabetes Association (ADA). The snack menu was recommended in accordance with the patient's favorites and conditions. The accuracy of the recommended menu was assessed in 2 steps. First, it was compared with the diet prescribed by three nutrition specialists. In the second step, system's suggested menu was evaluated by the data from 30 diabetic patients using a valid questionnaire. RESULTS: The results of evaluating the snack recommender system by nutritionists showed that this system is capable of recommending various snacks according to the season (accuracy of 100%) and personal interests (accuracy of 90%) to diabetic patients. According to health nutritionists, the snacks suggested by this system are matched with Iranian culture. Moreover, the results revealed that a higher body mass index (BMI) makes the recommender system less sensitive to personal interests to suggest what is basically beneficial for one's health. CONCLUSION: This study was a pioneering research to develop a more comprehensive dietary recommender system for diabetic patients which includes main meals as well. Patients found the system useful and were satisfied with the application. This system is believed to be able to help diabetic patients to take more healthy diet which leads to a better lifestyle.
BACKGROUND: As a prevalent metabolic disease, diabetes has different side effects and causes a wide range of co morbidity with a high rate of mortality. There is a need for certain interventions to manage this disease. Iranians usually have three main meals a day. Considering the special needs of diabeticpatients and the possibility of hypoglycemia between the main meals, it is essential for these patients to eat something as a snack. Considering these conditions and the society's orientation towards modern technologies such as smart phones, designing mobile-based nutrition recommender systems can be helpful. METHODS: The snack recommender system is a knowledge-based smart phone application. This study has focused on the development of a recommender system that combines artificial intelligence techniques and makes up a knowledge base according to the guidelines posed by the American Diabetes Association (ADA). The snack menu was recommended in accordance with the patient's favorites and conditions. The accuracy of the recommended menu was assessed in 2 steps. First, it was compared with the diet prescribed by three nutrition specialists. In the second step, system's suggested menu was evaluated by the data from 30 diabeticpatients using a valid questionnaire. RESULTS: The results of evaluating the snack recommender system by nutritionists showed that this system is capable of recommending various snacks according to the season (accuracy of 100%) and personal interests (accuracy of 90%) to diabeticpatients. According to health nutritionists, the snacks suggested by this system are matched with Iranian culture. Moreover, the results revealed that a higher body mass index (BMI) makes the recommender system less sensitive to personal interests to suggest what is basically beneficial for one's health. CONCLUSION: This study was a pioneering research to develop a more comprehensive dietary recommender system for diabeticpatients which includes main meals as well. Patients found the system useful and were satisfied with the application. This system is believed to be able to help diabeticpatients to take more healthy diet which leads to a better lifestyle.
Authors: S Abhari; R Safdari; L Azadbakht; K B Lankarani; Sh R Niakan Kalhori; B Honarvar; Kh Abhari; S M Ayyoubzadeh; Z Karbasi; S Zakerabasali; Y Jalilpiran Journal: J Biomed Phys Eng Date: 2019-12-01
Authors: Giang Thu Vu; Bach Xuan Tran; Roger S McIntyre; Hai Quang Pham; Hai Thanh Phan; Giang Hai Ha; Kenneth K Gwee; Carl A Latkin; Roger C M Ho; Cyrus S H Ho Journal: Int J Environ Res Public Health Date: 2020-03-17 Impact factor: 3.390