Recognition of multiple food, exercise and nutrition labels in glucoguide: phase IIa

葡萄糖指南中多种食品、运动和营养标签的识别：IIa 阶段

• 批准号: 486869-2015
• 负责人: Ling, Charles
• 金额: $ 9.11万
• 依托单位: University of Western Ontario
• 依托单位国家: 加拿大
• 项目类别: Idea to Innovation
• 财政年份: 2015
• 资助国家: 加拿大
• 起止时间: 2015-01-01 至 2016-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=585822
• 关键词: Recognition; multiple; food; exercise; nutrition

Diabetes is a chronic disease with many complications including heart attack, stroke, amputations and blindness. Self-management of blood glucose and improving lifestyle are very important to prevent complications, but it is a very difficult task. Mobile apps for diabetes tracking are increasingly available, but they can still be difficult to use - especially entering diet and exercise along with blood sugar and figuring out how to change diet and exercise to properly manage blood sugar. Therefore, we have been researching and developing GlucoGuide - an app that data-mines inputted diet and exercise data to give individual feedback and recommendations to help manage and reverse diabetes. This Phase IIa research and development project is a continuation of our Phase I I2I project (July 2014-July 2015) on GlucoGuide with promising results. Due to our achievements, we have secured early stage Angel Investment to our technologies, and the investor has put aside $70,000 specially to match this grant proposal. Continuing on our Phase I research and development, we will develop further analytic algorithms (including deep learning) to recognize food, multiple food combination, food nutrition label, barcode scanning and exercise equipment using the smartphone's camera. With simply taking a picture of meals, nutrition labels, and exercise equipment, users will be able to receive targeted recommendations to help manage their blood glucose better. This proposed research will significantly improve the usability of GlucoGuide, allowing for commercialization. The GlucoGuide Apps and system will benefit patients with pre-diabetes and diabetes, and overall, help to reduce the burden and cost on the healthcare systems in Canada and wrold.

糖尿病是一种慢性疾病，有许多并发症，包括心脏病发作，中风，截肢和失明。自我管理血糖和改善生活方式对预防并发症非常重要，但这是一项非常艰巨的任务。用于糖尿病跟踪的移动的应用程序越来越多，但它们仍然很难使用-特别是输入饮食和运动沿着血糖，并弄清楚如何改变饮食和运动以正确管理血糖。因此，我们一直在研究和开发GlucoGuide -一个应用程序，数据挖掘输入的饮食和运动数据，提供个人反馈和建议，以帮助管理和扭转糖尿病。 该IIa期研发项目是我们GlucoGuide I期I2 I项目（2014年7月至2015年7月）的延续，结果令人鼓舞。由于我们的成就，我们已经为我们的技术获得了早期天使投资，投资者已经专门拨出了70，000美元来匹配这项赠款提案。 继续我们的第一阶段研发，我们将开发进一步的分析算法（包括深度学习），以识别食物、多种食物组合、食物营养标签、条形码扫描和使用智能手机摄像头的运动器材。通过简单地拍摄膳食，营养标签和运动器材的照片，用户将能够获得有针对性的建议，以帮助更好地管理血糖。 这项拟议的研究将显著提高GlucoGuide的可用性，从而实现商业化。GlucoGuide应用程序和系统将使糖尿病前期和糖尿病患者受益，总体而言，有助于减轻加拿大和世界医疗保健系统的负担和成本。