Scholarlink

Integration of AI and geo-location technologies can transform healthcare services by providing more personalised and efficient recommendations. Current healthcare facility locators prioritise geographic proximity over specialised medical service. Due to reliance on static databases and needing more personalised, condition-centric matching, individuals need help finding suitable care based on specific conditions. Traditional systems often need to incorporate real-time data and advanced AI, limiting accuracy and usefulness. Recent advancements, particularly in GPT-3.5 and GPT-4, offer new opportunities for enhancing healthcare service recommendations. This study aims to develop a GeoAI-augmented web platform to recommend healthcare facilities for minor illnesses such as toothache, fever, headache, and stomachache. The platform combines AI-enhanced geo-location with precise condition-centric matching to improve user experience and accuracy, utilising a fine-tuned GPT-3.5 model to enhance interaction and generate relevant prompts. The system integrates OSM data to provide tailored facility suggestions. Methodologies include data collection of user prompts, model ne-tuning, expert reviews for relevance and accuracy, user surveys for satisfaction, and performance testing for response times. Comparative analysis between GPT-3.5 and GPT-4 assesses relevance, accuracy, user satisfaction, and response times. Results indicate that GPT-4 achieves higher accuracy and relevance, improved user satisfaction, and more contextually accurate responses compared to GPT-3.5, though with slightly longer response times. The study demonstrates GPT-4’s superior performance in generating accurate healthcare facility recommendations. The anticipated outcome is a chatbot-like web tool offering customised healthcare recommendations within specific regions, emphasising ease of use, precision, and contextual relevance. This study highlights the potential of AI and geo-location technologies in providing more personalised and effective healthcare solutions.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development and Evaluation of Geo-AI for Healthcare Discovery

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Myriame Virginie Matius,
  • Ivin Amri Musliman

摘要

Integration of AI and geo-location technologies can transform healthcare services by providing more personalised and efficient recommendations. Current healthcare facility locators prioritise geographic proximity over specialised medical service. Due to reliance on static databases and needing more personalised, condition-centric matching, individuals need help finding suitable care based on specific conditions. Traditional systems often need to incorporate real-time data and advanced AI, limiting accuracy and usefulness. Recent advancements, particularly in GPT-3.5 and GPT-4, offer new opportunities for enhancing healthcare service recommendations. This study aims to develop a GeoAI-augmented web platform to recommend healthcare facilities for minor illnesses such as toothache, fever, headache, and stomachache. The platform combines AI-enhanced geo-location with precise condition-centric matching to improve user experience and accuracy, utilising a fine-tuned GPT-3.5 model to enhance interaction and generate relevant prompts. The system integrates OSM data to provide tailored facility suggestions. Methodologies include data collection of user prompts, model ne-tuning, expert reviews for relevance and accuracy, user surveys for satisfaction, and performance testing for response times. Comparative analysis between GPT-3.5 and GPT-4 assesses relevance, accuracy, user satisfaction, and response times. Results indicate that GPT-4 achieves higher accuracy and relevance, improved user satisfaction, and more contextually accurate responses compared to GPT-3.5, though with slightly longer response times. The study demonstrates GPT-4’s superior performance in generating accurate healthcare facility recommendations. The anticipated outcome is a chatbot-like web tool offering customised healthcare recommendations within specific regions, emphasising ease of use, precision, and contextual relevance. This study highlights the potential of AI and geo-location technologies in providing more personalised and effective healthcare solutions.