|
审核状态: Project audit state: |
通过审核 Successful |
|
注册号: Registration number: |
ChiCTR2600119757 |
|
最近更新日期: Date of Last Refreshed on: |
2026-03-03 15:41:42 |
|
注册时间: Date of Registration: |
2026-03-03 00:00:00 |
|
注册号状态: |
预注册 |
|
Registration Status: |
Prospective registration |
|
注册题目: |
基于图文多模态大模型的超声心动图智能辅助系统效能评估 |
|
Public title: |
Performance evaluation of an intelligent echocardiography assistance system based on a multi-modal large model of text and images |
|
注册题目简写: |
|
|
English Acronym: |
|
|
研究课题的正式科学名称: |
多模态心脏超声大模型技术研发及智能诊疗体系构建研究 |
|
Scientific title: |
Research on the Development of Multimodal Cardiac Ultrasound Large Model Technology and the Construction of Intelligent Diagnosis and Treatment System |
|
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: |
卜安琪 |
研究负责人: |
尹磊 |
|
Applicant: |
Anqi Bu |
Study leader: |
Lei Yin |
|
申请注册联系人电话: Applicant telephone: |
+86 166 0372 2167 |
研究负责人电话:
Study leader's |
+86 136 0371 0217 |
|
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: |
baq981103@163.com |
研究负责人电子邮件: Study leader's E-mail: |
yinlei3008@163.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: |
郑州经开区经南十二路77号 |
研究负责人通讯地址: |
郑州经开区经南十二路77号 |
|
Applicant address: |
77 Jingnan 12th Rd, Jingkai District, Zhengzhou, Henan |
Study leader's address: |
77 Jingnan 12th Rd, Jingkai District, Zhengzhou, Henan |
|
申请注册联系人邮政编码: Applicant postcode: |
611418 |
研究负责人邮政编码: Study leader's postcode: |
611418 |
|
申请人所在单位: |
郑州市第七人民医院 |
||
|
Applicant's institution: |
The 7th People's Hospital of Zhengzhou |
||
|
研究负责人所在单位: |
郑州市第七人民医院 |
||
|
Affiliation of the Leader: |
The 7th People's Hospital of Zhengzhou |
||
|
是否获伦理委员会批准: |
是 |
||
|
Approved by ethic committee: |
Yes |
||
|
伦理委员会批件文号: Approved No. of ethic committee: |
科研伦审2025[ky-005] |
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: |
郑州市第七人民医院医学伦理委员会 |
||
|
Name of the ethic committee: |
Medical Ethics Committee of the 7th People's Hospital of Zhengzhou |
||
|
伦理委员会批准日期: Date of approved by ethic committee: |
2025-05-07 00:00:00 | ||
|
伦理委员会联系人: |
戴雅杰 |
||
|
Contact Name of the ethic committee: |
Yajie Dai |
||
|
伦理委员会联系地址: |
郑州经开区经南十二路77号 |
||
|
Contact Address of the ethic committee: |
77 Jingnan 12th Rd, Jingkai District, Zhengzhou, Henan |
||
|
伦理委员会联系人电话: Contact phone of the ethic committee: |
+86 135 9243 1714 |
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: |
郑州市第七人民医院 |
||||||||||||||||||||||
|
Primary sponsor: |
The 7th People's Hospital of Zhengzhou |
||||||||||||||||||||||
|
研究实施负责(组长)单位地址: |
郑州经开区经南十二路77号 |
||||||||||||||||||||||
|
Primary sponsor's address: |
77 Jingnan 12th Rd, Jingkai District, Zhengzhou, Henan |
||||||||||||||||||||||
|
试验主办单位(项目批准或申办者): Secondary sponsor: |
|
||||||||||||||||||||||
|
经费或物资来源: |
自筹 |
||||||||||||||||||||||
|
Source(s) of funding: |
Self-raised |
||||||||||||||||||||||
|
研究疾病: |
心血管疾病 |
||||||||||||||||||||||
|
Target disease: |
Cardiovascular diseases |
||||||||||||||||||||||
|
研究疾病代码: |
|
||||||||||||||||||||||
|
Target disease code: |
|
||||||||||||||||||||||
|
研究类型: |
诊断试验 |
||||||||||||||||||||||
|
Study type: |
Diagnostic test |
||||||||||||||||||||||
|
研究所处阶段: |
其它 | ||||||||||||||||||||||
|
Study phase: |
N/A |
||||||||||||||||||||||
|
研究设计: |
随机交叉对照 |
||||||||||||||||||||||
|
Study design: |
Cross-over |
||||||||||||||||||||||
|
研究目的: |
本研究不旨在验证AI替代医生,而是验证“医生+AI”的协同模式相比“传统手工”模式在以下两个维度的优越性: - 主要目的(提效): 验证AI辅助是否能显著缩短单例报告的周转时间(Turnaround Time)。 - 次要目的(增益): 验证AI辅助是否能提高(特别是低年资)医生对形态学异常的检出率(Sensitivity)。 |
||||||||||||||||||||||
|
Objectives of Study: |
This study does not aim to verify the replacement of doctors by AI, but rather to examine the superiority of the "doctor +AI" collaborative model over the "traditional manual" model in the following two dimensions: - Main objective (efficiency improvement) : To verify whether AI-assisted assistance can significantly reduce the Turnaround Time of single-case reports. Secondary objective (gain) : To verify whether AI-assisted assistance can enhance the detection rate (Sensitivity) of morphological abnormalities by (especially junior) doctors. |
||||||||||||||||||||||
|
药物成份或治疗方案详述: |
|
||||||||||||||||||||||
|
Description for medicine or protocol of treatment in detail: |
|
||||||||||||||||||||||
|
纳入标准: |
本研究的研究对象包含两个层面:1. 受试医生;2. 评估所用病例(图像)。 1. 受试医生标准 (1)持有有效的医师执业证书,执业范围包含医学影像诊断。 (2)根据从业经验分为两组: ???? 低年资组:独立从事超声心动图诊断及报告书写工作 ≤ 5年。 ???? 高年资组:独立从事超声心动图诊断及报告书写工作 ≥ 10年,且具备主治医师或以上职称。 (3)当前每周常规完成超声心动图报告不少于10例。 (4)自愿参加并签署书面知情同意书。 (5)能够熟练操作计算机,并保证研究期间可访问院内部署的研究系统。 2. 评估病例(图像)标准 用于本次研究的超声心动图静态图像及动态循环需满足: (1)图像质量合格,能够满足临床诊断的基本要求(心腔结构、瓣膜等关键区域显示清晰)。 (2)图像包含常规切面(如胸骨旁左室长轴、短轴、心尖四腔心、两腔心等)及必要的测量数据。 (3)病例诊断覆盖范围广,需包含: ???? 正常或无明显结构异常的病例(占比约30-40%)。 ???? 具有明确形态学异常的病例(占比约60-70%),异常类型须涵盖本研究重点关注的类型。 |
||||||||||||||||||||||
|
Inclusion criteria |
The research subjects of this study consist of two levels: 1. The participating doctors; 2. Evaluate the cases (images) used. 1. Criteria for the test physician (1) Hold a valid physician's practice certificate, with the scope of practice including medical imaging diagnosis. (2) Divided into two groups based on professional experience: ? Junior group: Independently engaged in echocardiographic diagnosis and report writing for no more than 5 years. ? Senior Group: Independently engaged in echocardiographic diagnosis and report writing for at least 10 years, and holding the title of attending physician or above. (3) Currently, no less than 10 echocardiogram reports are routinely completed every week. (4) Participate voluntarily and sign a written informed consent form. (5) Be proficient in operating computers and ensure access to the research systems deployed within the institute during the research period. 2. Criteria for evaluating cases (images) The static images and dynamic cycles of echocardiography used in this study must meet the following requirements: (1) The image quality is qualified and can meet the basic requirements for clinical diagnosis (clear display of key areas such as heart cavity structure and valves). (2) The images include conventional sections (such as the long axis, short axis, apical four-chamber heart, two-chamber heart, etc. of the left ventricle beside the sternum) and necessary measurement data. (3) The case diagnosis should have a wide coverage and include: Cases that are normal or have no obvious structural abnormalities |
||||||||||||||||||||||
|
排除标准: |
1. 受试医生排除标准 符合以下任一条件的医生予以排除: (1)非超声心动图专业或日常工作中不包含心脏超声诊断的医师。 (2)直接参与本研究验证AI系统的算法开发或训练数据标注的人员。 (3)与研究项目资助方存在可能影响结果客观性的直接利益冲突。 (4)计划在研究期间长期外出,无法完成全部评估病例者。 2. 评估病例(图像)排除标准 符合以下任一条件的病例图像不予采用: (1)图像质量极差,无法进行可靠的视觉评估。 (2)患者信息脱敏不彻底,存在隐私泄露风险。 (3)病例诊断不明确,即便经专家委员会审议仍无法达成共识确定“金标准”。 |
||||||||||||||||||||||
|
Exclusion criteria: |
Exclusion criteria for the participating doctors Doctors who meet any of the following conditions will be excluded: (1) Physicians who are not specialized in echocardiography or whose daily work does not involve cardiac ultrasound diagnosis. (2) Personnel who directly participated in the algorithm development or training data annotation of the AI system verified in this study. (3) There is a direct conflict of interest with the research project funder that may affect the objectivity of the results. (4) Those who plan to be away for a long time during the research period and are unable to complete all the evaluated cases. 2. Evaluate the exclusion criteria for cases (images) Case images that meet any of the following conditions will not be accepted: (1) The image quality is extremely poor, making it impossible to conduct reliable visual assessment. (2) The patient information is not desensitized thoroughly, posing a risk of privacy leakage. (3) The diagnosis of the case is not clear. Even after review by the expert committee, no consensus can be reached to determine the "gold standard". |
||||||||||||||||||||||
|
研究实施时间: Study execute time: |
从 From 2026-03-09 00:00:00至 To 2027-12-31 00:00:00 |
征募观察对象时间: Recruiting time: |
从 From 2026-03-09 00:00:00 至 To 2027-12-31 00:00:00 |
|
诊断试验: Diagnostic Tests: |
|
||||||||||||||||||||||||||||
|
研究实施地点: Countries of recruitment and research settings: |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
测量指标: Outcomes: |
|
|
采集人体标本:
Collecting sample(s)
|
|
|
征募研究对象情况: Recruiting status: |
尚未开始 Not yet recruiting |
年龄范围: Participant age: |
|
||||||
|
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): |
对所有报告进行统一去标识化处理,隐去报告来源、医师或AI标识、阶段信息,仅保留统一的随机评价编号。将报告根据随机化种子数随机排序,确保同一病例不同来源的报告不会连续出现。 |
||||||||
|
Randomization Procedure (please state who generates the random number sequence and by what method): |
All the reports were deidentified, with the source of the report, the identification of the doctor or AI, and the stage information hidden, and only the unified random evaluation number was retained. The reports were randomly ordered according to the number of randomization seeds to ensure that reports from different sources of the same case did not appear consecutively. |
||||||||
|
是否公开试验完成后的统计结果: Calculated Results after the Study Completed public access: |
不公开/Private |
|
盲法: |
对受试者设盲(单盲):受试医生(读者)在开始每个具体病例的判读前,不知晓自己被分配使用的是传统模式还是AI辅助模式。这是通过上述的分配隐藏和技术屏蔽实现的。 对结局评估者设盲:负责评估主要结局(报告周转时间)和次要结局(将医生报告与专家参考标准进行比对、进行报告质量评分)的研究人员,在进行分析时,对病例所属的判读模式(分组)不知情。报告数据将以盲态提交给评估者。 |
|
Blinding: |
Blinding of participants (readers): The physician-readers are blinded to the allocation (traditional vs. AI-assisted mode) before starting the interpretation of each case, through allocation concealment and technical masking. Blinding of outcome assessors: Personnel analyzing the primary outcome (turnaround time) and secondary outcomes (agreement with reference standard, report quality scoring) are blinded to the group allocation during assessment. |
|
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): |
无 |
|
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: |
一、病例记录表: 1. 病例信息:病例编号、图像来源、专家委员会裁定的金标准诊断。 2. 过程数据:医生编号与分组、评估阶段(传统/AI辅助模式)、系统自动记录的报告开始时间与提交时间。 二、EDC: 1. 采集方式:关键的操作时间与文本数据均由研究系统后台自动记录,最大限度减少人工录入错误。 2. 系统部署:采用基于医院内部网络部署的专用数据采集与管理平台,所有数据存储于医院服务器,确保数据不出院,保障信息安全与患者隐私。 3. 数据管理:所有病例数据均进行脱敏处理。 |
|
Data collection and Management (A standard data collection and management system include a CRF and an electronic data capture: |
I. Case Record Form: Case information: Case number, image source, and the gold standard diagnosis determined by the expert committee. 2. Process data: Doctor number and grouping, assessment stage (traditional/AI-assisted mode), start time and submission time of reports automatically recorded by the system. Ii. EDC 1. Collection method: Key operation times and text data are automatically recorded by the research system's backend, minimizing manual entry errors to the greatest extent. 2. System Deployment: A dedicated data collection and management platform based on the hospital's internal network is adopted. All data is stored on the hospital server to ensure that no data is discharged, safeguarding information security and patient privacy. 3. Data Management: All case data are desensitized. |
|
数据与安全监察委员会: Data and Safety Monitoring Committee: |
有/Yes |