利用多种可解释性机器学习模型预测胆囊占位性病变良恶性的多中心回顾性研究

注册号:

Registration number:

ChiCTR2600121723 

最近更新日期:

Date of Last Refreshed on:

2026-04-02 08:40:25 

注册时间:

Date of Registration:

2026-04-02 00:00:00 

注册号状态:

预注册

Registration Status:

Prospective registration

注册题目:

利用多种可解释性机器学习模型预测胆囊占位性病变良恶性的多中心回顾性研究

Public title:

Prediction of Benign and Malignant Gallbladder Space-Occupying Lesions Using Interpretable Machine Learning Models: A Multicenter Retrospective Study

注册题目简写:

English Acronym:

研究课题的正式科学名称:

利用多种可解释性机器学习模型预测胆囊占位性病变良恶性的多中心回顾性研究

Scientific title:

Prediction of Benign and Malignant Gallbladder Space-Occupying Lesions Using Interpretable Machine Learning Models: A Multicenter Retrospective Study

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:

彭程 

研究负责人:

彭程 

Applicant:

Peng Cheng 

Study leader:

Peng Cheng 

申请注册联系人电话:

Applicant telephone:

+86 18560082089

研究负责人电话:

Study leader's
telephone:

+86 531 82166651

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:

Dr.Peng@email.sdu.edu.cn

研究负责人电子邮件:

Study leader's E-mail:

Dr.Peng@email.sdu.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:

中国山东省济南市历下区文化西路107号

研究负责人通讯地址:

中国山东省济南市历下区文化西路107号

Applicant address:

No. 107, Wenhua West Road, Lixia District, Jinan, Shandong, China

Study leader's address:

No. 107, Wenhua West Road, Lixia District, Jinan, Shandong, China

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:

山东大学齐鲁医院

Applicant's institution:

Qilu Hospital of Shandong University

研究负责人所在单位:

山东大学齐鲁医院

Affiliation of the Leader:

Qilu Hospital of Shandong University

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:

KYLL-2026-02-004

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:

山东大学齐鲁医院科研伦理委员会

Name of the ethic committee:

Scientific Research Ethics Committee of Qilu Hospital of Shandong University

伦理委员会批准日期:

Date of approved by ethic committee:

2026-02-09 00:00:00

伦理委员会联系人:

卜丽娟

Contact Name of the ethic committee:

Bu Lijuan

伦理委员会联系地址:

中国山东省济南市历下区文化西路107号

Contact Address of the ethic committee:

No. 107, Wenhua West Road, Lixia District, Jinan, Shandong, China

伦理委员会联系人电话:

Contact phone of the ethic committee:

+86 531 82169166

伦理委员会联系人邮箱:

Contact email of the ethic committee:

bulijuan16@sdu.edu.cn

研究实施负责(组长)单位:

山东大学齐鲁医院

Primary sponsor:

Qilu Hospital of Shandong University

研究实施负责(组长)单位地址:

中国山东省济南市历下区文化西路107号

Primary sponsor's address:

No. 107, Wenhua West Road, Lixia District, Jinan, Shandong, China

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

市(区县):

Country:

China

Province:

Shandong

City:

单位(医院):

山东大学齐鲁医院

具体地址:

中国山东省济南市历下区文化西路107号

Institution
hospital:

Qilu Hospital of Shandong University

Address:

No. 107, Wenhua West Road, Lixia District, Jinan, Shandong, China

经费或物资来源:

自选课题(自筹)

Source(s) of funding:

Self-selected topic (self-funded)

研究疾病:

胆囊占位性病变和胆囊癌  

Target disease:

Gallbladder space-occupying lesions and Gallbladder cancer

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:

系统比较多种机器学习算法,整合多中心临床数据,开发并验证一个高性能、可解释的机器学习模型,用于术前精准预测胆囊占位性病变的良恶性,为临床制定个体化诊疗决策提供可靠工具。  

Objectives of Study:

This study systematically compared multiple machine learning algorithms and integrated multicenter clinical data to develop and validate a high-performance, interpretable machine learning model for the preoperative prediction of benign and malignant gallbladder space-occupying lesions, thereby providing a reliable tool to support individualized clinical decision-making.

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.临床病历资料缺失严重,关键预测变量缺失率超过50%;
2.术后病理诊断不明确;
3.既往有胆囊手术史;
4.就诊前患有其它恶性疾病;

Exclusion criteria:

1.Patients with severely missing clinical data, where the missing rate of key predictor variables exceeded 50%, were excluded.
2.Postoperative pathological diagnosis was unclear / inconclusive.
3.History of previous gallbladder surgery.
4.Presence of other malignant diseases prior to presentation.

研究实施时间:

Study execute time:

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

征募观察对象时间:

Recruiting time:

From 2026-04-10 00:00:00 To 2026-08-31 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):

术后病理诊断

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):

The final diagnosis was confirmed by postoperative pathology.

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):

待评价的诊断方法为本研究旨在开发的可解释机器学习预测模型,它们分别是C5.0决策树、高斯过程(GP)、逻辑回归(LR)、多层感知器(MLP)、朴素贝叶斯(NB)、神经网络(NN)、XGBoost(XGB)

Index test:

The interpretable machine learning models developed in this study—namely, C5.0、GP、LR、MLP、NB、NN、XGB—were evaluated as diagnostic tools.

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)

目标人群是胆囊癌患者

例数:

Sample size:

200

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):

The target population is patients with gallbladder cancer.

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):

胆囊占位性病变为胆囊腺肌症、较大的炎性息肉或胆固醇性息肉等良性病变的患者

例数:

Sample size:

1200

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

Patients with benign gallbladder space-occupying lesions, such as adenomyomatosis, large inflammatory polyps, or cholesterol polyps.

研究实施地点:

Countries of recruitment and research settings:

国家:

中国

省(直辖市):

山东 

市(区县):

 

Country:

China

Province:

Shandong

City:

单位(医院):

山东大学齐鲁医院 

单位级别:

三级甲等 

Institution
hospital:

Qilu Hospital of Shandong University

Level of the institution:

Tertiary A

国家:

中国

省(直辖市):

山东 

市(区县):

 

Country:

China

Province:

Shandong

City:

单位(医院):

山东省立医院 

单位级别:

三级甲等 

Institution
hospital:

Shandong Provincial Hospital

Level of the institution:

Tertiary A

国家:

中国

省(直辖市):

山东 

市(区县):

 

Country:

China

Province:

Shandong

City:

单位(医院):

山东省千佛山医院 

单位级别:

三级甲等 

Institution
hospital:

Shandong Provincial Qianfoshan Hospital

Level of the institution:

Tertiary A

测量指标:

Outcomes:

指标中文名:

受试者工作特征曲线下面积

指标类型:

主要指标

Outcome:

Area under the receiver operating characteristic curve (AUC)

Type:

Primary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

以假阳性率(1-特异度)为横坐标、真阳性率(敏感度)为纵坐标绘制 ROC 曲线,计算曲线下面积。

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

Plot the ROC curve with the false positive rate (1 - specificity) as the x-axis and the true positive rate (sensitivity) as the y-axis, and calculate the area under the curve.

指标中文名:

决策曲线分析

指标类型:

次要指标

Outcome:

Decision Curve Analysis(DCA)

Type:

Secondary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

指标中文名:

敏感度

指标类型:

次要指标

Outcome:

Sensitivity

Type:

Secondary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

真阴性 /(真阴性 + 假阳性)

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

True Negative / (True Negative + False Positive)

指标中文名:

特异度

指标类型:

次要指标

Outcome:

Specificity

Type:

Secondary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

真阴性 /(真阴性 + 假阳性)

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

True Negative / (True Negative + False Positive)

指标中文名:

F1分数

指标类型:

次要指标

Outcome:

F1 score

Type:

Secondary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

2 ×(PPV × 敏感度)/(PPV + 敏感度)

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

(2 × (PPV × Sensitivity)) / (PPV + Sensitivity)

指标中文名:

准确率

指标类型:

次要指标

Outcome:

Accuracy

Type:

Secondary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

(真阳性 + 真阴性)/ 总样本数

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

(True positive + True negative) / Total sample number

指标中文名:

阳性预测值

指标类型:

次要指标

Outcome:

Positive predictive value(PPV)

Type:

Secondary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

真阳性 /(真阳性 + 假阳性)

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

True Positive / (True Positive + False Positive)

指标中文名:

阴性预测值

指标类型:

次要指标

Outcome:

Negative predictive value(NPV)

Type:

Secondary indicator

测量时间点:

模型训练完成后,分别在训练集、内部测试集、外部测试集1、外部测试集2上进行评估

测量方法:

真阴性 /(真阴性 + 假阴性)

Measure time point of outcome:

After the model training is completed, it will be evaluated respectively on the training set, intern

Measure method:

True Negative / (True Negative + False Negative)

采集人体标本:

Collecting sample(s)
from participants:

标本中文名:

组织:

Sample Name:

N/A

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 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):

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:

通过病例记录表采集数据,并储存于电子管理系统中

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

Data were collected using a Case Report Form (CRF) and stored in an electronic data management system.

数据与安全监察委员会:

Data and Safety Monitoring Committee:

有/Yes

注册人:

Name of Registration:

 2026-04-02 08:40:18