大模型在阴道镜报告质控与解读中的应用研究

注册号:

Registration number:

ChiCTR2500114886 

最近更新日期:

Date of Last Refreshed on:

2025-12-18 15:03:51 

注册时间:

Date of Registration:

2025-12-18 00:00:00 

注册号状态:

预注册

Registration Status:

Prospective registration

注册题目:

大模型在阴道镜报告质控与解读中的应用研究

Public title:

Research on the Application of Large Models in the Quality Control and Interpretation of Colposcopy Reports

注册题目简写:

English Acronym:

研究课题的正式科学名称:

Colpo-LLM辅助阴道镜报告质控与解读的模型开发验证与随机对照研究

Scientific title:

Development, validation and randomized controlled study of a model for quality control and interpretation of colposcopy reports assisted by Colpo-LLM

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:

王宇婷 

研究负责人:

薛鹏;代倩苓 

Applicant:

Yuting Wang 

Study leader:

Peng Xue; Qianling Dai 

申请注册联系人电话:

Applicant telephone:

+86 181 0489 7753

研究负责人电话:

Study leader's
telephone:

+86 150 1109 4125

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:

wangyuting_0112@163.com

研究负责人电子邮件:

Study leader's E-mail:

xuepeng_pumc@foxmail.com

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:

北京市东城区北极阁三条31号院

研究负责人通讯地址:

北京市东城区北极阁三条31号院 成都市青羊区日月大道1617号

Applicant address:

No. 31, Beijige Santiao, Dongcheng District, Beijing

Study leader's address:

No. 31, Beijige Santiao, Dongcheng District, Beijing; 1617 Riyue Avenue, Qingyang District, Chengdu City.

申请注册联系人邮政编码:

Applicant postcode:

100730

研究负责人邮政编码:

Study leader's postcode:

100730; 610073

申请人所在单位:

中国医学科学院北京协和医学院群医学及公共卫生学院

Applicant's institution:

School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College

研究负责人所在单位:

中国医学科学院北京协和医学院 成都市妇女儿童中心医院

Affiliation of the Leader:

Chinese Academy of Medical Sciences and Peking Union Medical College; chengdu Women's and Children's central 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:

CAMS&PUMC-IEC-2025-048

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:

中国医学科学院北京协和医学院涉及人的生物医学研究伦理审查委员会

Name of the ethic committee:

The Ethics Review Committee for Biomedical Research Involving Human Subjects of the Chinese Academy of Medical Sciences and Peking Union Medical College

伦理委员会批准日期:

Date of approved by ethic committee:

2025-03-14 00:00:00

伦理委员会联系人:

雷晶

Contact Name of the ethic committee:

Jing Lei

伦理委员会联系地址:

北京市东城区东单三条9号院

Contact Address of the ethic committee:

No. 9, Dongdan Third Street, Dongcheng District, Beijing

伦理委员会联系人电话:

Contact phone of the ethic committee:

+86 10 6510 5894

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:

中国医学科学院北京协和医学院 成都市妇女儿童中心医院

Primary sponsor:

Chinese Academy of Medical Sciences and Peking Union Medical College; chengdu Women's and Children's central Hospital

研究实施负责(组长)单位地址:

北京市东城区北极阁三条31号院 成都市青羊区日月大道1617号

Primary sponsor's address:

No. 31, Beijige Santiao, Dongcheng District, Beijing; 1617 Riyue Avenue, Qingyang District, Chengdu City.

试验主办单位(项目批准或申办者):

Secondary sponsor:

国家:

中国

省(直辖市):

北京市

市(区县):

Country:

China

Province:

Beijing

City:

单位(医院):

中国医学科学院北京协和医学院群医学及公共卫生学院

具体地址:

北京市东城区北极阁三条31号院

Institution
hospital:

School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College

Address:

No. 31, Beijige Santiao, Dongcheng District, Beijing

经费或物资来源:

人社部中国博士后科学基金[GZB20230076]

Source(s) of funding:

Ministry of Human Resources and Social Security - China Postdoctoral Science Foundation

研究疾病:

宫颈癌  

Target disease:

Cervical Cancer

研究疾病代码:

Target disease code:

研究类型:

干预性研究

Study type:

Interventional study

研究所处阶段:

其它 

Study phase:

N/A

研究设计:

随机平行对照 

Study design:

Parallel 

研究目的:

本研究基于已开发的大语言模型Colpo-LLM,聚焦于阴道镜报告的两个关键应用方向:医生端的报告质量控制与患者端的报告内容解读。通过随机对照试验(RCT)设计,全面评估该系统在提升报告规范性、诊断准确性和患者理解水平方面的实际效能。具体目标如下: 1.规范阴道镜报告的质量控制,提升报告标准化水平 通过开发基于大模型的自动质控系统,实现对阴道镜报告的实时检测,及时发现报告中的错误、遗漏和不规范之处,并发出预警,提醒医生进行修改和完善以提高阴道镜报告的文书质量,推动阴道镜诊疗行为的规范化、同质化和标准化,减少人为差异。 2.提升阴道镜报告解读的准确性与患者参与感 通过开发精准且通俗易懂的阴道镜报告解读系统,将专业术语转化为易于理解的语言,帮助患者更快、更准确地理解阴道镜检查结果,减少因等待和信息不对称而引发的焦虑情绪。提升患者对自己健康状况的认知和参与感,促使患者主动参与诊疗决策,进而提高其依从性及治疗效果。 3.推动智能化医疗技术的应用与普及 本研究通过智能化技术在阴道镜报告质量控制与解读支持中的应用,为智能化技术在其他医学检查领域的推广应用提供了有力的支持和实践依据,推动了医疗智能化的进一步发展。  

Objectives of Study:

This study is based on the already developed large language model Colpo-LLM, focusing on two key application areas of colposcopy reports: quality control of reports on the doctor’s side and content interpretation of reports on the patient’s side. Through a randomized controlled trial (RCT) design, the study comprehensively evaluates the system's effectiveness in improving report standardization, diagnostic accuracy, and patient comprehension. The specific objectives are as follows: 1. Standardize colposcopy report quality control and improve report standardization By developing an automatic quality control system based on a large model, this study aims to achieve real-time detection of colposcopy reports, promptly identifying errors, omissions, and inconsistencies in the reports. The system will issue alerts to remind doctors to make necessary revisions and improvements to enhance the quality of the reports, promoting the standardization, homogenization, and normalization of colposcopy diagnostic practices, and reducing human variation. 2. Improve the accuracy of colposcopy report interpretation and patient engagement By developing an accurate and easy-to-understand colposcopy report interpretation system, professional terminology will be converted into language that patients can easily comprehend, helping them to quickly and accurately understand the results of their colposcopy examination. This will reduce anxiety caused by waiting and information asymmetry. The system will enhance patients' understanding of their health status and involvement in their healthcare decisions, leading to increased treatment adherence and improved outcomes. 3. Promote the application and popularization of intelligent healthcare technology This study, by applying intelligent technology in colposcopy report quality control and interpretation support, provides strong support and practical evidence for the promotion and application of intelligent technologies in other medical examination fields, driving further development in healthcare automation.

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. 正在接受心理治疗或重度焦虑状态; 2. 不具有阴道镜检查指征已知晓宫颈癌的病理确诊结果; 3. 有严重认知障碍或精神疾病; 4. 无法完整完成试验流程者。

Exclusion criteria:

1. Currently undergoing psychotherapy or in a state of severe anxiety; 2. Without indications for colposcopy and already aware of the pathological diagnosis of cervical cancer; 3. Suffering from severe cognitive impairment or mental illness; 4. Unable to complete the entire trial process.

研究实施时间:

Study execute time:

From 2025-12-24 00:00:00 To 2026-06-08 00:00:00  

征募观察对象时间:

Recruiting time:

From 2025-12-24 00:00:00 To 2026-06-08 00:00:00

干预措施:

Interventions:

组别:

干预组——患者

样本量:

257

Group:

Intervention group——patients

Sample size:

干预措施:

给患者同步发放模型生成的患者版解读文本

干预措施代码:

Intervention:

Simultaneously distribute the patient version interpretation text generated by the model to the patients.

Intervention code:

组别:

干预组——阴道镜医生

样本量:

257

Group:

Intervention group——colposcopists

Sample size:

干预措施:

医生参考Colpo-LLM质控辅助输出

干预措施代码:

Intervention:

The colposcopists referred to the Colpo-LLM quality control auxiliary output.

Intervention code:

研究实施地点:

Countries of recruitment and research settings:

国家:

中国

省(直辖市):

四川省 

市(区县):

成都 

Country:

China

Province:

Sichuan Province

City:

Chengdu

单位(医院):

成都市妇女儿童中心医院 

单位级别:

三甲 

Institution
hospital:

Chengdu Women's and Children's Central Hospital

Level of the institution:

Tertiary A

国家:

中国

省(直辖市):

内蒙古自治区 

市(区县):

鄂尔多斯市 

Country:

China

Province:

Inner Mongolia Autonomous Region

City:

Erdos

单位(医院):

鄂尔多斯市妇幼保健院 

单位级别:

二级 

Institution
hospital:

Erdos Maternal and Child HealthCare Hospital

Level of the institution:

Secondary

国家:

中国

省(直辖市):

江苏省 

市(区县):

无锡市 

Country:

China

Province:

Jiangsu Province

City:

Wuxi

单位(医院):

无锡市妇幼保健院 

单位级别:

三甲 

Institution
hospital:

Wuxi Maternal and Child HealthCare Hospital

Level of the institution:

Tertiary A

测量指标:

Outcomes:

指标中文名:

个性化客观理解测试分数(解读)

指标类型:

主要指标

Outcome:

Personalized Objective Understanding Test Score (Interpretation)

Type:

Primary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

焦虑评分变化(解读)

指标类型:

次要指标

Outcome:

Anxiety Score Changes (Interpretation)

Type:

Secondary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

主观理解评分(解读)

指标类型:

次要指标

Outcome:

Subjective Understanding Score (Interpretation)

Type:

Secondary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

主观评价得分(解读)

指标类型:

次要指标

Outcome:

Subjective evaluation score (Interpretation)

Type:

Secondary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

灵敏度(质控)

指标类型:

主要指标

Outcome:

Sensitivity (Quality Control)

Type:

Primary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

特异度(质控)

指标类型:

主要指标

Outcome:

Specificity (Quality Control)

Type:

Primary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

准确率(质控)

指标类型:

次要指标

Outcome:

Accuracy (Quality Control)

Type:

Secondary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

F1值(质控)

指标类型:

次要指标

Outcome:

F1 Score (Quality Control)

Type:

Secondary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

处理效率(质控)

指标类型:

次要指标

Outcome:

Processing Efficiency (Quality Control)

Type:

Secondary indicator

测量时间点:

测量方法:

Measure time point of outcome:

Measure method:

指标中文名:

任务依从性(质控)

指标类型:

次要指标

Outcome:

Task Compliance (Quality Control)

Type:

Secondary 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

其它  

说明

Fate of sample:

0thers  

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 18 years
最大 Max age 64 years

性别:

女性

Gender:

Female

随机方法(请说明由何人用什么方法产生随机序列):

本研究采用分层随机化(Stratified Randomization)方法。根据3个临床中心和3种病变类型(NILM、LSIL、HSIL)作为分层因素,共构建9个独立的随机化层(3中心 × 3病变类型)。 随机分配序列由一名独立于临床诊断及受试者招募的统计人员生成,使用 Python 编程语言中的 random 模块完成。为确保随机过程的可重复性,已设定并保存随机数种子(Random Seed)。在每一个分层内,受试者依据生成的序列独立随机分配至干预组(A组)或对照组(B组),直至达到该分层的预定样本量。

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

The study utilizes a stratified randomization method based on clinical centers (3 sites) and cervical lesion types (NILM, LSIL, and HSIL), resulting in a total of 9 independent strata. The randomization sequences were generated by an independent statistician who is not involved in clinical diagnosis or subject recruitment, using the random module in Python. A random seed has been saved to ensure the reproducibility of the sequence. Within each stratum, subjects are independently randomized into either the Intervention Group (A) or the Control Group (B) until the target sample size for that stratum is reached.

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:

诊断盲法:阴道镜医生仅负责完成临床的拟诊(NILM、LSIL、HSIL),不掌握受试者的分组信息。即,医生不知晓患者被分配到干预组还是对照组。这确保了在分组过程中的盲法执行,从而避免了医生的偏见或期望对诊断结果的影响。分组操作由现场研究协调员根据系统自动提示进行,确保诊断医生无法获知任何关于分组的信息

Blinding:

Diagnostic blinding: The colposcopist is only responsible for completing the clinical diagnosis (NILM, LSIL, HSIL) and does not have access to the participant's group allocation information. In other words, the doctor is unaware of whether the patient has been assigned to the intervention or control group. This ensures blinding during the allocation process, preventing any potential bias or expectations from the doctor influencing the diagnostic results. The allocation process is carried out by the on-site research coordinator based on automated system prompts, ensuring that the diagnosing physician does not have access to any information about the group allocation.

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:

1.病例记录表: 在本研究中,患者接受阴道镜检查后,医生将生成并打印电子阴道镜报告。报告内容包括患者的基本信息(如年龄)、HPV检测结果、细胞学检查结果、阴道镜所见以及其他相关的临床数据。该电子报告将作为病例记录表(CRF)的一部分,由经过专门培训的阴道镜医生进行填写,确保数据的准确性和完整性。在填写过程中,研究人员将严格遵循研究协议和标准操作程序(SOP),以保证研究数据的一致性和规范性。此外,电子阴道镜系统所配套的病例管理系统将保存患者的所有相关信息,并进行实时更新。这些信息将有助于确保数据的完整性,并提供清晰、可靠的参考记录供后续研究分析。 2.电子数据采集和管理系统(EDC): 为确保数据采集和管理的高效性和准确性,我们将采用电子阴道镜系统的病例管理系统以及医院的HIS系统进行数据的采集、存储和管理。在整个研究过程中,所有相关问卷将通过问卷星平台进行设计、分发和收集。通过这一平台,我们可以实时录入、跟踪并管理研究数据,确保数据的完整性、准确性以及安全性。该系统具备自动数据核对和异常检查功能,可以有效减少人为错误的发生,进一步确保数据质量。此外,为了保护参与者的隐私,所有采集到的数据将进行加密存储,并且定期进行数据备份,以防止数据丢失。所有操作都严格遵循数据保护法律法规,确保研究过程中的数据隐私得到充分保障。通过这种电子化管理方式,我们可以大幅提高数据采集的效率和可靠性,确保研究数据的可追溯性,并且为后续数据分析提供坚实的基础。

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

1.Case Record Form (CRF): In this study, after the patients undergo colposcopy, the doctor will generate and print an electronic colposcopy report. The report will include the patient's basic information (such as age), HPV test results, cytology findings, colposcopic observations, and other relevant clinical data. This electronic report will serve as part of the Case Record Form (CRF) and will be filled out by a specially trained colposcopist to ensure the accuracy and completeness of the data. During the filling process, researchers will strictly adhere to the study protocol and Standard Operating Procedures (SOP) to ensure the consistency and standardization of the data. In addition, the case management system associated with the electronic colposcopy system will store all related patient information and provide real-time updates. This information will help ensure the integrity of the data and provide a clear, reliable reference record for subsequent research analysis. 2.Electronic Data Capture and Management System(EDC): To ensure the efficiency and accuracy of data collection and management, we will use the case management system associated with the electronic colposcopy system and the hospital's HIS system for data capture, storage, and management. Throughout the study, all relevant questionnaires will be designed, distributed, and collected through the Wenjuanxing platform. This platform will allow us to input, track, and manage research data in real-time, ensuring the integrity, accuracy, and security of the data. The system is equipped with automatic data validation and anomaly detection features, which can effectively reduce human error and further ensure data quality. Furthermore, to protect participants' privacy, all collected data will be encrypted and stored, with regular backups to prevent data loss. All operations will strictly comply with data protection laws and regulations, ensuring the full protection of data privacy throughout the research process. Through this electronic management approach, we can significantly enhance the efficiency and reliability of data collection, ensure the traceability of research data, and provide a solid foundation for subsequent data analysis.

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:

 2025-12-18 15:03:46