论文部分内容阅读
目的:研发针对常见眼病的智能问答系统，并调查和评估其在互联网医院平台的应用效果，为医疗服务模式和技术创新提供借鉴。方法:利用自然语言处理技术研发常见眼病智能问答系统，通过对预先录入的结构化知识进行解析分词和关键词采集绑定等处理，经过搜索引擎为用户提供准确的问答内容。并通过功能、性能测试，完成互联网医院的部署。收集2020年2月1—29日期间使用常见眼病智能问答系统的数据资料，对患者的人口学信息、问诊时间、咨询疾病类别、咨询内容、服务评价等进行分析。结果:在中山大学中山眼科中心互联网医院成功研发出具有智能导诊服务和眼病术后智能咨询模块的常见眼病智能问答系统，测试结果表明智能问答系统的可重复性为100.0%，准确性为99.8%，可24 h持续在线服务。2020年2月1—29日期间中山眼科中心互联网医院常见眼病智能问答系统总服务患者6 462人次，其中男3 082人次，占47.7%，女3 380人次，占52.3%。患者的平均年龄为32.3岁。1 135例患者使用了智能导诊服务，占17.6%，5 375例患者使用了眼病术后智能咨询模块，占82.4%。平均每日使用常见眼病智能问答系统的患者为223人次，最高每小时服务74人次。分别有25.6%(38/148)和36.4%(54/148)的患者对常见眼病智能问答系统的服务效率表示非常满意和比较满意；24.3%(36/148)和37.8%(56/148)的患者对常见眼病智能问答系统的服务质量分别表示非常满意和比较满意。结论:常见眼病智能问答系统具有较高的可重复性和准确性。基于互联网医院平台的智能问答系统应用可满足患者24 h医疗咨询需求，降低医院交叉感染风险和医务人员的工作压力。“,”Objective:To survey the application of an intelligent consultation system for common eye diseases and evaluate its applicational effectiveness on an internet hospital platform.Methods:A cross-sectional study was performed in Zhongshan Ophthalmic Centre of Sun Yat-sen University.Natural language processing technology was applied to develop the intelligent consultation system for common eye diseases.Its efficiency and quality were evaluated.The survey data were collected from February 1 to 29, 2020 to analyze the demographic information, consultation time, consultation category, consultation content, service satisfaction.This study protocal was approved by an Ethic Committee of Zhongshan Ophthalmic Centre of Sun Yat-sen University(2020KYPJ095).Results:The intelligent consultation system for common eye diseases was developed and successfully deployed in Internet Hospital of Zhongshan Ophthalmic Center.The repeatability and accuracy of the intelligent consultation system were 100.0% and 99.8%, respectively.During February 1 to 29, 2020, the intelligent consultation system served 6 462 patients, including 3 082 males(47.7%) and 3 380 females(52.3%). The average age of patients was 32.3 years old.Total of 1 135(17.6%) patients used the intelligent guidance consultation, and 5 375(82.4%) patients used the intelligent outpatient consultation.The intelligence consultation system was applied by 223 patients per day with a maximum of 74 patients per hour.The survey showed that 25.6% and 36.4% of the patients felt very satisfied and relatively satisfied with the efficiency of the intelligent consultation service, respectively; 24.3% and 37.8% of the patients were very satisfied and relatively satisfied with the quality of the intelligent consultation service, respectively.Conclusions:Intelligent consultation system for common eye diseases can meet the needs of patients because of its high repeatability and accuracy.Patients are satisfied with the service efficiency and quality of the intelligent consultation system, which avoids the risk of cross infection and releases the burden of medical staff.