ChiCTR2500115797 版本V1.0 版本创建时间2025/12/31 11:11:44 中国临床试验注册中心

审核状态:

Project audit state:

通过审核

Successful

注册号:

Registration number:

ChiCTR2500115797 

最近更新日期:

Date of Last Refreshed on:

2025-12-31 11:11:39 

注册时间:

Date of Registration:

2025-12-31 00:00:00 

注册号状态:

预注册

Registration Status:

Prospective registration

注册题目:

脊柱侧弯多模态AI辅助诊断与生物力学分析平台

Public title:

Evaluating the Diagnostic Accuracy of a Multimodal AI Platform for Scoliosis: A Retrospective Cohort Study on Classification, Cobb Angle Measurement, and Biomechanical Analysis

注册题目简写:

English Acronym:

研究课题的正式科学名称:

脊柱侧弯多模态AI辅助诊断与生物力学分析平台

Scientific title:

A Multimodal AI-Assisted Diagnostic and Biomechanical Analysis Platform for Scoliosis

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:

赵妍 

研究负责人:

鲍圣涌 

Applicant:

Zhao Yan 

Study leader:

Bao Shengyong 

申请注册联系人电话:

Applicant telephone:

+86 755 22942613

研究负责人电话:

Study leader's
telephone:

+86 755 22943054

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:

214012310@qq.com

研究负责人电子邮件:

Study leader's E-mail:

271334972@qq.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:

广东省深圳市罗湖区东门北路1017号

研究负责人通讯地址:

深圳市罗湖区东门北路1017号

Applicant address:

NO.1017, Dongmen North Road, Luohu District, Shenzhen, Guangdong Province, China

Study leader's address:

Dongmen North Road 1017, Luohu, Shenzhen

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:

深圳市人民医院

Applicant's institution:

Shenzhen People’s Hospital

研究负责人所在单位:

深圳市人民医院

Affiliation of the Leader:

Shenzhen People's 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:

LL-KY-2025283-01

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:

深圳市人民医院科研伦理委员会

Name of the ethic committee:

Research Ethics Committee of ShenZhen People's Hospital

伦理委员会批准日期:

Date of approved by ethic committee:

2025-11-21 00:00:00

伦理委员会联系人:

骆瑜

Contact Name of the ethic committee:

LuoYu

伦理委员会联系地址:

深圳市罗湖区东门北路1017号

Contact Address of the ethic committee:

Dongmen North Road 1017, Luohu, Shenzhen

伦理委员会联系人电话:

Contact phone of the ethic committee:

+86 755 22943881

伦理委员会联系人邮箱:

Contact email of the ethic committee:

195323995@qq.com

研究实施负责(组长)单位:

深圳市人民医院

Primary sponsor:

Shenzhen People's Hospital

研究实施负责(组长)单位地址:

深圳市罗湖区东门北路1017号

Primary sponsor's address:

Dongmen North Road 1017, Luohu, Shenzhen

试验主办单位(项目批准或申办者):

Secondary sponsor:

国家:

中国

省(直辖市):

广东省

市(区县):

Country:

China

Province:

Guangdong

City:

单位(医院):

深圳市人民医院

具体地址:

深圳市罗湖区东门北路1017号

Institution
hospital:

Shenzhen People's Hospital

Address:

Dongmen North Road 1017, Luohu, Shenzhen

经费或物资来源:

自选课题(自筹)

Source(s) of funding:

Investigator-Initiated

研究疾病:

脊柱侧弯  

Target disease:

Scoliosis

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:

诊断试验诊断准确性 

Study design:

Diagnostic test for accuracy 

研究目的:

主要目的:验证多模态AI平台在脊柱侧弯诊断中的准确性,具体包括:1. 七级分类功能与专家共识金标准相比的诊断效能(AUC);2. 测量Cobb角的精确性(平均绝对误差, MAE)。 次要目的:1.评估AI性能在不同人口学亚组中的一致性;2.探索AI生物力学分析结果与有限元模型的一致性;3.将AI的诊断性能与单名临床专家的初始评估进行比较。  

Objectives of Study:

Primary Objective: To validate the diagnostic accuracy of a multimodal AI platform for scoliosis, specifically including: 1) the diagnostic performance (AUC) of its seven-level classification function compared to the expert consensus gold standard; and 2) the precision of its Cobb angle measurement (Mean Absolute Error, MAE).Secondary Objectives: 1) To evaluate the consistency of AI performance across different demographic subgroups; 2) To explore the agreement between AI biomechanical analysis results and those from finite element models; 3) To compare the diagnostic performance of the AI with the initial assessment of a single clinical expert.

药物成份或治疗方案详述:

 

Description for medicine or protocol of treatment in detail:

 

纳入标准:

1.符合诊断标准的特发性脊柱侧弯患者;
2.具备符合质量要求的站立位脊柱全景正侧位X光片(DICOM格式或PNG格式且分辨率不低于2K);
3.电子病历系统中具备完整的临床记录和影像学报告;

Inclusion criteria

1.Patients diagnosed with idiopathic scoliosis.
2.Availability of standing whole-spine anteroposterior and lateral radiographs meeting quality requirements (in DICOM format, or PNG format with a resolution not less than 2K).
3.Complete clinical records and radiology reports available in the electronic medical record system.

排除标准:

1.其他原因引起的脊柱侧弯患者(非特发性脊柱侧弯);
2.既往有脊柱外科手术史者;
3.关键影像学资料质量不合格,存在严重运动伪影、图像模糊或椎体关键解剖结构缺失,影响判读和测量者;
4.有限元仿真失败者;

Exclusion criteria:

1.Patients with scoliosis caused by other etiologies (non-idiopathic scoliosis).
2.Patients with a history of spinal surgery.
3.Patients with key imaging data of poor quality, including severe motion artifacts, blurring, or missing key vertebral anatomical structures that affect interpretation and measurement.
4.Cases where finite element simulation fails.

研究实施时间:

Study execute time:

From 2026-01-01 00:00:00 To 2027-12-31 00:00:00  

征募观察对象时间:

Recruiting time:

From 2026-01-01 00:00:00 To 2026-12-31 00:00:00

诊断试验:

Diagnostic Tests:

金标准或参考标准(即可准确诊断某疾病的单项方法或多项联合方法,在本研究中用于诊断是否有该病的临床参考标准):

金标准(七级分类与Cobb角):由两名高年资(副主任医师及以上)放射科或脊柱外科医生组成独立评审小组。在互不知晓对方结果和AI结果的情况下(双盲),根据患者的全部影像学资料(X光、CT)和临床信息,参照SRS-Schwab成人脊柱畸形分类等标准,独立完成七级分类判定和Cobb角手工测量。两位专家的初始评估结果将被单独记录。当两人判定不一致时,将由第三名更高年资的专家进行仲裁,以最终达成的一致意见作为最终金标准。

Gold Standard or Reference Standard (The clinical reference standards required to establish the presence or absence of the target condition in the tested population in present study):

Gold Standard (Seven-level Classification and Cobb Angle): An independent review panel consisting of two senior radiologists or spine surgeons (holding the rank of associate chief physician or higher) will be formed. In a double-blind manner (unaware of each other's results and the AI outputs), the

指标试验(即本研究的待评估诊断试验,无论为方法、生物标志物或设备,均请列出名称):

AI脊柱侧弯七级分类模型、AI Cobb角自动测量模型、AI脊柱生物力学分析模型

Index test:

an AI Scoliosis Seven-Level Classification Model, an AI Cobb Angle Automated Measurement Model, an AI Spinal Biomechanical Analysis Model.

目标人群(可以是某种疾病患者或正常人群,详细描述其疾病特征,注意应纳入符合分布特点的全序列病例,具有良好的代表性)

2018-2024年深圳市人民医院确诊的、年龄10-80岁的特发性脊柱侧弯(Cobb角≥10°)患者。

例数:

Sample size:

800

Target condition (The target condition is a particular disease or disease stage that the index test will be intended to identify. Please specify the characteristics in detail; the population should has a complete spectrum and good representative):

Patients diagnosed with idiopathic scoliosis (Cobb angle ≥10°) at Shenzhen People's Hospital between 2018 and 2024, aged 10 to 80 years.

容易混淆的疾病人群(即与目标疾病不易区分的一种或多种不同疾病,应避免采用正常人群对照的病例-对照设计):

例数:

Sample size:

0

Population with condition difficult to distinguish from the target condition, the normal population in a case-control study design should be avoid:

None

研究实施地点:

Countries of recruitment and research settings:

国家:

中国

省(直辖市):

广东省 

市(区县):

 

Country:

China

Province:

Guangdong

City:

单位(医院):

深圳市人民医院 

单位级别:

三级甲等 

Institution
hospital:

Shenzhen People's Hospital

Level of the institution:

Tertiary A

测量指标:

Outcomes:

指标中文名:

生物力学指标的ICC

指标类型:

次要指标

Outcome:

ICC for Biomechanical Indicators

Type:

Secondary indicator

测量时间点:

基于回顾性数据的一次性评估

测量方法:

在具备配对CT数据的亚组中,计算AI生物力学模型输出的应力指标与高保真有限元模型(参考标准) 计算的对应指标之间的ICC。

Measure time point of outcome:

Evaluation is based on a single assessment of retrospective data.

Measure method:

Within the subgroup possessing paired CT data, the Intraclass Correlation Coefficient (ICC) will be calculated between the stress indicators output by the AI biomechanical model and the corresponding indicators calculated by the high-fidelity finite element model (the reference standard).

指标中文名:

两位专家初始评估间的一致性

指标类型:

次要指标

Outcome:

Agreement Between Initial Assessments by Two Experts

Type:

Secondary indicator

测量时间点:

基于回顾性数据的一次性评估

测量方法:

分类:计算专家A初始分类与专家B初始分类之间的线性加权Kappa值。 连续:计算专家A初始Cobb角测量与专家B初始Cobb角测量之间的ICC。

Measure time point of outcome:

Evaluation is based on a single assessment of retrospective data.

Measure method:

Categorical Variable: The linearly weighted Kappa value will be calculated between the initial classification by Expert A and the initial classification by Expert B. Continuous Variable: The Intraclass Correlation Coefficient (ICC) will be calculated between the initial Cobb angle measurement by Expert A and the initial measurement by Expert B.

指标中文名:

七级分类的AUC

指标类型:

主要指标

Outcome:

AUC for Seven-level Classification

Type:

Primary indicator

测量时间点:

基于回顾性数据的一次性评估

测量方法:

将AI平台输出的1-7级分类结果,与金标准(两位专家仲裁后的一致意见) 进行比较,计算AUC及其95%置信区间。

Measure time point of outcome:

Evaluation is based on a single assessment of retrospective data.

Measure method:

The Area Under the Curve (AUC) and its 95% confidence interval will be calculated by comparing the 1-7 level classification results output by the AI platform with the gold standard (the consensus opinion reached after arbitration by two experts).

指标中文名:

Cobb角测量的ICC

指标类型:

次要指标

Outcome:

ICC for Cobb Angle Measurement

Type:

Secondary indicator

测量时间点:

基于回顾性数据的一次性评估

测量方法:

分别计算AI测量值、专家A初始测量值、专家B初始测量值 各自与 金标准测量值 之间的ICC(双向随机,绝对一致性)。

Measure time point of outcome:

Evaluation is based on a single assessment of retrospective data.

Measure method:

The Intraclass Correlation Coefficient (ICC) (two-way random, absolute agreement) will be calculated separately for the AI-measured value, Expert A's initial measurement, and Expert B's initial measurement, each against the gold standard measurement.

指标中文名:

Cobb角测量的MAE

指标类型:

主要指标

Outcome:

MAE for Cobb Angle Measurement

Type:

Primary indicator

测量时间点:

基于回顾性数据的一次性评估

测量方法:

计算AI自动测量的Cobb角与金标准(两位专家仲裁后的一致手工测量值) 之间差值的绝对值的平均数。

Measure time point of outcome:

Evaluation is based on a single assessment of retrospective data.

Measure method:

The Mean Absolute Error (MAE) will be calculated as the average of the absolute differences between the Cobb angles automatically measured by the AI and the gold standard (the consensus manual measurement value after arbitration by two experts).

指标中文名:

AI与单名专家初始评估的一致性

指标类型:

次要指标

Outcome:

Agreement Between AI and a Single Expert's Initial Assessment

Type:

Secondary indicator

测量时间点:

基于回顾性数据的一次性评估

测量方法:

分类:计算AI与专家A初始分类、AI与专家B初始分类之间的线性加权Kappa值。 连续:计算AI与专家A初始Cobb角测量、AI与专家B初始Cobb角测量之间的ICC。

Measure time point of outcome:

Evaluation is based on a single assessment of retrospective data.

Measure method:

Categorical Variable: The linearly weighted Kappa value will be calculated between the AI classification and the initial classification by Expert A, and between the AI classification and the initial classification by Expert B. Continuous Variable: The Intraclass Correlation Coefficient (ICC) will be calculated between the AI Cobb angle measurement and the initial measurement by Expert A, and between the AI measurement and the initial measurement by Expert B.

指标中文名:

七级分类的灵敏度、特异度

指标类型:

次要指标

Outcome:

Sensitivity and Specificity for Seven-level Classification

Type:

Secondary indicator

测量时间点:

基于回顾性数据的一次性评估

测量方法:

将七级分类按“一对一”方式转化为二分类问题,分别计算AI分类结果相对于金标准在每个等级上的灵敏度和特异度。

Measure time point of outcome:

Evaluation is based on a single assessment of retrospective data.

Measure method:

The seven-level classification results will be transformed into binary classification problems using a "one-vs-rest" approach. The sensitivity and specificity of the AI classification results will be calculated for each level against the gold standard.

采集人体标本:

Collecting sample(s)
from participants:

标本中文名:

组织:

Sample Name:

NA

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 10 years
最大 Max age 80 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):

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

None

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:

Blinding:

None

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):

不适用

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

N/A

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:

数据采集:经培训的研究员使用标准化电子病例报告表(e-CRF)从PACS、HIS和EMR系统中提取去标识化数据。金标准标注由专家在盲态下独立完成。 数据管理:SOP培训:所有数据采集和标注人员均需通过标准化操作流程培训。盲法设计:专家标注与AI分析相互盲;统计分析人员对分组盲。数据核查:设置逻辑核查(如Cobb角范围核查);随机抽取10%数据进行二次复核。代码开源:本研究的统计分析代码将在发表后于GitHub平台开源,以促进可重复性。

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

Data Collection: Trained researchers will extract de-identified data from the PACS, HIS, and EMR systems using a standardized electronic case report form (e-CRF). The gold standard annotations will be independently completed by experts under blinded conditions.Data Management:SOP Training: All personnel involved in data collection and annotation must complete standardized operating procedure (SOP) training.Blinding Design: Expert annotation and AI analysis are mutually blinded; statistical analysts are blinded to group assignments.Data Verification: Logical checks (e.g., Cobb angle range verification) are implemented, and 10% of the data is randomly selected for secondary review.Code Open-Source: The statistical analysis code for this study will be made publicly available on the GitHub platform after publication to promote reproducibility.

数据与安全监察委员会:

Data and Safety Monitoring Committee:

无/No

注册人:

Name of Registration:

 2025-12-31 11:11:39