大型语言模型在真实世界冠心病诊断中的评价

注册号:

Registration number:

ChiCTR2600123358 

最近更新日期:

Date of Last Refreshed on:

2026-04-25 00:33:48 

注册时间:

Date of Registration:

2026-04-25 00:00:00 

注册号状态:

预注册

Registration Status:

Prospective registration

注册题目:

大型语言模型在真实世界冠心病诊断中的评价

Public title:

Evaluation of Large Language Models in Real World Coronary Artery Disease Diagnosis

注册题目简写:

English Acronym:

研究课题的正式科学名称:

大型语言模型在真实世界冠心病诊断中的评价

Scientific title:

Evaluation of Large Language Models in Real World Coronary Artery Disease Diagnosis

研究课题代号(代码):

Study subject ID:

在二级注册机构或其它机构的注册号:

The registration number of the Partner Registry or other register:

申请注册联系人:

纪宏伟 

研究负责人:

郭俊杰 

Applicant:

Ji Hongwei 

Study leader:

Guo Junjie 

申请注册联系人电话:

Applicant telephone:

+86 131 2051 8791

研究负责人电话:

Study leader's
telephone:

+86 186 6180 5902

申请注册联系人传真 :

Applicant Fax:

研究负责人传真:

Study leader's fax:

申请注册联系人电子邮件:

Applicant E-mail:

hongweijicn@gmail.com

研究负责人电子邮件:

Study leader's E-mail:

guojunjie@qdu.edu.cn

申请单位网址(自愿提供):

Applicant website(voluntary supply):

研究负责人网址(自愿提供):

Study leader's website(voluntary supply):

申请注册联系人通讯地址:

浙江省杭州市解放路88号

研究负责人通讯地址:

山东省青岛市市南区江苏路16号

Applicant address:

No. 88 Jiefang Road, Hangzhou City, Zhejiang Province

Study leader's address:

No.16 Jiangsu Road, Shinan District, Qingdao City, Shandong Province

申请注册联系人邮政编码:

Applicant postcode:

研究负责人邮政编码:

Study leader's postcode:

申请人所在单位:

浙江大学医学院附属第二医院

Applicant's institution:

The Second Affiliated Hospital of Zhejiang University School of Medicine

研究负责人所在单位:

青岛大学附属医院

Affiliation of the Leader:

Qingdao University Affiliated Hospital

是否获伦理委员会批准:

Approved by ethic committee:

Yes

伦理委员会批件文号:

Approved No. of ethic committee:

[临]伦 QYFYEC2025-30-01

伦理委员会批件附件:

Approved file of Ethical Committee:

查看附件View

批准本研究的伦理委员会名称:

青岛大学附属医院医学伦理委员会

Name of the ethic committee:

Medical Ethics Committee of Qingdao University Affiliated Hospital

伦理委员会批准日期:

Date of approved by ethic committee:

2026-03-18 00:00:00

伦理委员会联系人:

朱婕

Contact Name of the ethic committee:

Zhu Jie

伦理委员会联系地址:

青岛大学附属医院行政楼

Contact Address of the ethic committee:

Administrative Building of Qingdao University Affiliated Hospital

伦理委员会联系人电话:

Contact phone of the ethic committee:

+86 532 8291 1869

伦理委员会联系人邮箱:

Contact email of the ethic committee:

研究实施负责(组长)单位:

青岛大学附属医院

Primary sponsor:

The Affiliated Hospital of Qingdao University

研究实施负责(组长)单位地址:

山东省青岛市江苏路16号, 266000

Primary sponsor's address:

No.16 Jiangsu Road, Qingdao City, Shandong Province, 266000

试验主办单位(项目批准或申办者):

Secondary sponsor:

国家:

中国

省(直辖市):

山东省

市(区县):

Country:

China

Province:

Shandong

City:

单位(医院):

青岛大学附属医院

具体地址:

山东省青岛市江苏路16号, 266000

Institution
hospital:

The Affiliated Hospital of Qingdao University

Address:

No.16 Jiangsu Road, Qingdao City, Shandong Province, 266000

经费或物资来源:

国家自然科学基金

Source(s) of funding:

National Natural Science Foundation of China

研究疾病:

冠心病  

Target disease:

Coronary Artery Disease

研究疾病代码:

Target disease code:

研究类型:

观察性研究

Study type:

Observational study

研究所处阶段:

其它 

Study phase:

N/A

研究设计:

队列研究 

Study design:

Cohort study 

研究目的:

本研究为探索性卫生服务研究,旨在评估一种基于大型语言模型(LLM)的辅助诊断工具在冠心病诊疗决策中的潜在价值与决策合适性等医疗质量/效率。具体目标包括: 1. 开发基于通义千问的微调模型,用于从非侵入性入院数据中预测阻塞性冠心病风险; 2. 在回顾性数据中,进行数据再分析,比较LLM辅助下医生与纯医生判断冠脉造影(ICA)必要性的决策合适率(同时符合造影金标准与AUC指南)等医疗质量和效率,不干预临床决策,属于卫生服务实施前评估。  

Objectives of Study:

This study is an exploratory health service research aimed at evaluating the potential value and suitability of a large language model (LLM) based diagnostic tool in coronary heart disease diagnosis and treatment decisions, as well as medical quality/efficiency. The specific goals include: 1. Developing a fine-tuning model based on the universal question to predict the risk of obstructive coronary heart disease from non-invasive admission data; 2. In retrospective data, conducting data reanalysis to compare the appropriateness of decision-making between doctors and pure doctors in determining the necessity of coronary angiography (ICA) with LLM assistance (while complying with the gold standard for angiography and AUC guidelines) and other medical quality and efficiency, without interfering with clinical decision-making, which belongs to pre-implementation evaluation of health services.

药物成份或治疗方案详述:

 

Description for medicine or protocol of treatment in detail:

 

纳入标准:

Inclusion criteria

排除标准:

1. 既往确诊冠心病(心肌梗死史/血运重建史); 2. 以急性冠脉综合征(ACS)就诊; 3. 临床资料不完整; 4. 心电图无法评估(起搏器心律/预激综合征)。

Exclusion criteria:

Prior diagnosis of coronary artery disease (history of myocardial infarction or prior revascularization) Presentation with acute coronary syndrome (ACS) Incomplete pre‑angiography clinical data ECG not interpretable (e.g., pacemaker rhythm, pre‑excitation syndrome)

研究实施时间:

Study execute time:

From 2026-04-30 00:00:00 To 2026-06-30 00:00:00  

征募观察对象时间:

Recruiting time:

From 2026-04-30 00:00:00 To 2026-06-30 00:00:00

干预措施:

Interventions:

组别:

纯医生判断组

样本量:

500

Group:

Physician only group

Sample size:

干预措施:

无(仅提供临床资料)

干预措施代码:

Intervention:

None (only clinical information provided)

Intervention code:

组别:

LLM辅助判断组

样本量:

500

Group:

LLM assisted group

Sample size:

干预措施:

无,提供临床资料 + LLM预测结果

干预措施代码:

Intervention:

None. Clinical information + LLM prediction results provided

Intervention code:

研究实施地点:

Countries of recruitment and research settings:

国家:

中国

省(直辖市):

山东省 

市(区县):

 

Country:

China

Province:

Shandong

City:

单位(医院):

青岛大学附属医院 

单位级别:

三甲 

Institution
hospital:

The Affiliated Hospital of Qingdao University

Level of the institution:

Tertiary A

测量指标:

Outcomes:

指标中文名:

决策合适率

指标类型:

主要指标

Outcome:

Decision appropriateness

Type:

Primary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

采集人体标本:

Collecting sample(s)
from participants:

标本中文名:

组织:

Sample Name:

none

Tissue:

人体标本去向

使用后销毁  

说明

Fate of sample:

Destruction after use  

Note:

征募研究对象情况:

Recruiting status:

尚未开始

Not yet recruiting

年龄范围:

Participant age:

最小 Min age 30 years
最大 Max age years

性别:

男女均可

Gender:

Both

随机方法(请说明由何人用什么方法产生随机序列):

本研究为回顾性数据的再评估。依据Nature Medicine 32, 616–623 (2026)和 Nature 616, 520–524 (2023)等研究,在回顾性数据中对病例进行随机化评估顺序,将由主要研究者使用计算机生成的随机数字表对病例序列进行随机排序后分配给不同评估组。

Randomization Procedure (please state who generates the random number sequence and by what method):

This study is a re-evaluation of retrospective data. According to studies such as Nature Medicine 32, 616-623 (2026) and Nature 616, 520-524 (2023), in retrospective cases, cases were randomized and evaluated. The main researcher used a computer-generated random number table to randomly sort the case sequence and assign it to different evaluation group.

是否公开试验完成后的统计结果:

Calculated Results after the Study Completed public access:

公开/Public

盲法:

本研究在回顾性数据评估过程中对独立结果评估者施盲,评估者不知道病例所属的评估组别(纯医生组 vs LLM辅助组)。

Blinding:

In the evaluation process of retrospective data, independent outcome assessors were blinded and did not know the evaluation group to which the case belonged (pure physician group vs LLM assisted group).

试验完成后的统计结果(上传文件):

Calculated Results after
the Study Completed(upload file):

是否共享原始数据:

IPD sharing

否No

共享原始数据的方式(说明:请填入公开原始数据日期和方式,如采用网络平台,需填该网络平台名称和网址):

The way of sharing IPD”(include metadata and protocol, If use web-based public database, please provide the url):

None

数据采集和管理(说明:数据采集和管理由两部分组成,一为病例记录表(Case Record Form, CRF),二为电子采集和管理系统(Electronic Data Capture, EDC),如ResMan即为一种基于互联网的EDC:

电子采集和管理系统。本研究的数据采集和管理由独立研究者(非临床诊疗团队)负责所有数据在采集时即进行去标识化处理,不包含患者姓名、身份证号、住院号等个人可识别信息。独立研究者无权访问任何患者身份信息,仅处理匿名化临床变量。

Data collection and Management (A standard data collection and management system include a CRF and an electronic data capture:

Electronic Data Capture, EDC. The independent researcher (non clinical diagnosis and treatment team) is responsible for the data collection and management of this study. All data will be de identified at the time of collection, excluding the patient's name, ID number, hospitalization number and other personally identifiable information. Independent researchers have no access to any patient identification information and only handle anonymized clinical variables.

数据与安全监察委员会:

Data and Safety Monitoring Committee:

暂未确定/Not yet

注册人:

Name of Registration:

 2026-04-25 00:33:43