聊天机器人对有关勃起功能障碍最常见问题的回答质量。
Quality of Chatbot Responses to the Most Popular Questions Regarding Erectile Dysfunction.
作者信息
Barlas İrfan Şafak, Tunç Lütfi
机构信息
Clinic of Urology, Ankara Acibadem Hospital, Ankara, Türkiye.
出版信息
Urol Res Pract. 2025 Jan 3;50(4):253-260. doi: 10.5152/tud.2025.24098.
OBJECTIVE
Erectile dysfunction (ED) is a common cause of male sexual dysfunction. We aimed to evaluate the quality of ChatGPT and Gemini's responses to the most frequently asked questions about ED.
METHODS
This study was conducted as a cross-sectional, observational study. Google Trends was used to determine the most frequently asked questions on the internet. ChatGPT-3.5 and Gemini were compared for these chatbots' answers to the questions about ED. Two urologists with board certificates assessed the quality of responses using the Global Quality Score (GQS).
RESULTS
Fifteen questions about ED were included according to the Google Trends. ChatGPT was able to answer all the questions systematically, whereas Gemini could not answer two questions. Upon assessing the quality of the responses provided by both researchers with the GQS, it was observed that the frequency of low-quality responses from Gemini exceeded that of ChatGPT. The agreement between researchers was 92% for ChatGPT and 95% for Gemini.
CONCLUSION
Despite the expeditious and comprehensive answers provided by chatbots, we identified inadequacies in their responses related to ED. In their current state, they cannot replace the patient-centered approach of healthcare professionals and require further development.
目的
勃起功能障碍（ED）是男性性功能障碍的常见原因。我们旨在评估ChatGPT和Gemini对有关ED的最常见问题的回答质量。
方法
本研究作为一项横断面观察性研究进行。利用谷歌趋势来确定互联网上最常见的问题。比较了ChatGPT-3.5和Gemini对有关ED问题的回答。两名获得委员会认证的泌尿科医生使用全球质量评分（GQS）评估回答的质量。
结果
根据谷歌趋势，纳入了15个有关ED的问题。ChatGPT能够系统地回答所有问题，而Gemini无法回答两个问题。在用GQS评估两位研究人员提供的回答质量时，发现Gemini低质量回答的频率超过了ChatGPT。研究人员之间对ChatGPT的一致性为92%，对Gemini为95%。
结论
尽管聊天机器人提供了迅速而全面的回答，但我们发现它们有关ED的回答存在不足之处。就目前的状态而言，它们无法取代以患者为中心的医疗专业人员的方法，需要进一步发展。
Zotero 插件