以公衛三段五級策略設計與開發精準營養聊天機器人

Design and develop a precision nutrition chatbot using the public health three-stage five-level strategy

目標：回應當前社會面臨的兩大挑戰：非傳染性疾病（NCDs）的擴散和全球人口老齡化趨勢。本研究因應公衛三段五級的策略設計與開發一款精準營養聊天機器人，以幫助用戶更好地控制和管理民眾的健康飲食並對抗NCDs。此外，本研究之聊天機器人運用現代行動科技特性，具備社群網路基礎下更高效、人性化和可持續的照顧系統，以應對人口老齡化帶來的照顧人力不足的挑戰。方法：以精準營養的概念為基礎，公衛三段五級照護為策略，以Line 社群為平台，建立可以完成營養三段五級健康照顧的精準營養聊天機器人。結果：本研究建立之精準營養聊天機器人，使用QRcode 即可透過Line 社群平台擴及使用者。進入聊天機器人後，首先，初段預防，民眾可以使用線上課程功能，觀看到營養師所設計與設立的營養知識內容，達到初級健康促進。再者，次級篩檢，配合營養篩檢目的，建立篩檢推播訊息，讓需要進行營養篩檢之民眾，透過手機即可即時進行營養狀況篩檢，例如：營養不良篩檢、衰弱狀況篩檢等。最後，末段營養照護，營養狀況篩檢後，確立營養狀況不佳者，可使用與營養師諮詢的功能，上傳飲食內容，營養師可以透過後台，分析營養成分，系統可自動計算營養素以及熱量，並且給予建議。結論：本研究設計了一款建置於Line 通訊平台上的精準營養聊天機器人，透過社群網站即可快速擴及，針對台灣長者群體，有效地利用他們對該平台的熟悉度，來實施三段五級的精準營養照護計劃。

Objective: To respond to the two major challenges faced by current society: the spread of non-communicable diseases (NCDs) and the global trend of population aging. This study develops aprecision nutrition chatbot in response to public health's three-tier five-level strategy, aiming to help users’better control and manage their healthy diet and combat NCDs. In addition, this chatbot utilizes modern mobile technology features, providing amore efficient, humanized, and sustainable care system under social network infrastructure to address the challenge of insufficient caregiving manpower due to population aging. Method: Based on the concept of precision nutrition and using public health's three-tier five- level care as astrategy, we establish aprecision nutrition chatbot that can complete three-tier five-level nutritional healthcare on Line community platform. Result: The precision nutrition chatbot developed in this study can be accessed by users through QR code via Line community platform. After entering the chatbot, firstly for primary prevention, people can use online course function to view nutritional knowledge content designed by dietitians for basic health promotion. Secondly for secondary screening purpose, we set up push messages for screening so that those who need nutritional screening can conduct real-time assessment on their phone such as malnutrition or frailty condition screenings etc.. Finally, for end-stage nutritional care, after screening the nutritional status and identifying those with poor nutrition, they can use the function to consult with adietitian. They can upload their dietary content, and the dietitian can analyze the nutritional components through the backend. The system can automatically calculate nutrients and calories and provide suggestions. Conclusion: This study designed aprecise nutrition chatbot built on the LINE communication platform, which can be quickly expanded through social networking sites. It effectively utilizes the familiarity of Taiwanese elderly with this platform to implement athree-stage five-level precision nutrition care plan.