ChiCTR2600120951 版本V1.0 版本创建时间2026/03/23 17:16:18 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：

 ChiCTR2600120951 

最近更新日期：

Date of Last Refreshed on：

 2026-03-23 17:16:13 

注册时间：

Date of Registration：

 2026-03-23 00:00:00 

注册号状态：

预注册

Registration Status：

Prospective registration

注册题目：

基于深度学习的实体瘤智能诊断与疗效评估的多中心研究

Public title：

Deep Learning-Based Multicenter Study on Intelligent Diagnosis and Treatment Response Assessment of Solid Tumors

注册题目简写：

English Acronym：

研究课题的正式科学名称：

基于深度学习的实体瘤智能诊断与疗效评估的多中心研究

Scientific title：

Deep Learning-Based Multicenter Study on Intelligent Diagnosis and Treatment Response Assessment of Solid Tumors

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：

胡晓云 

研究负责人：

胡晓云 

Applicant：

XiaoYun Hu 

Study leader：

Xiaoyun Hu 

申请注册联系人电话：

Applicant telephone：

 +86 510 82700775

研究负责人电话：

Study leader's
telephone：

+86 510 85350345

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：

 drxyh@foxmail.com

研究负责人电子邮件：

Study leader's E-mail：

 drxyh@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：

无锡市清扬路299号

研究负责人通讯地址：

无锡市清扬路299号

Applicant address：

No. 299 Qingyang Road, Wuxi City

Study leader's address：

Wuxi Qingyang road 299

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：

南京医科大学附属无锡人民医院

Applicant's institution：

The Affiliated Wuxi People's Hospital of Nanjing Medical University

研究负责人所在单位：

无锡市人民医院

Affiliation of the Leader：

Wuxi people’s Hospital

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：

 KY26030

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：

无锡市人民医院科研伦理委员会

Name of the ethic committee：

Research Ethics Committee of Wuxi People's Hospital

伦理委员会批准日期：

Date of approved by ethic committee：

 2026-02-03 00:00:00

伦理委员会联系人：

彭雁

Contact Name of the ethic committee：

Peng Yan

伦理委员会联系地址：

无锡市清扬路299号

Contact Address of the ethic committee：

Wuxi Qingyang road 299

伦理委员会联系人电话：

Contact phone of the ethic committee：

 +86 510 85350835

伦理委员会联系人邮箱：

Contact email of the ethic committee：

 76489926@qq.com

研究实施负责（组长）单位：

无锡市人民医院

Primary sponsor：

Wuxi people’s Hospital

研究实施负责（组长）单位地址：

无锡市清扬路299号

Primary sponsor's address：

Wuxi Qingyang road 299

试验主办单位(项目批准或申办者)：

Secondary sponsor：

国家：

中国

省(直辖市)：

江苏省

市(区县)：

Country：

China

Province：

Jiangsu

City：

单位(医院)：

无锡市人民医院

具体地址：

无锡市清扬路299号

Institution
hospital：

Wuxi people’s Hospital

Address：

Wuxi Qingyang road 299

经费或物资来源：

“雁阵人才”学科带头人-胡晓云-2025

Source(s) of funding：

Wild Goose Formation Talent Program: Discipline Leader - Hu Xiaoyun - 2025

研究疾病：

实体瘤 ；非小细胞肺癌，症状主要包括持续性咳嗽、痰中带血（咯血）、胸痛、呼吸困难、声音嘶哑以及不明原因的体重下降与乏力。  

Target disease：

Solid tumors; Non-small cell lung cancer (NSCLC), symptoms mainly include persistent cough, hemoptysis (blood in sputum), chest pain, shortness of breath, hoarseness, and unexplained weight loss and fatigue.

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：

 探索性研究/预试验 

Study phase：

0

研究设计：

队列研究 

Study design：

Cohort study 

研究目的：

本研究旨在解决非小细胞肺癌（NSCLC）免疫化学治疗中疗效评估滞后及“假性进展”误判等临床难题。拟构建多中心、多时序标准化数据库，利用时空图卷积网络（ST-GCN）与分解-集成 Transformer 模型，深度融合多时序 CT 影像及临床特征，动态量化肿瘤微环境的时空异质性演变。最终确立“动态 iRECIST”智能评估标准，开发具备高可解释性的自动化评估系统，实现疗效的早期、精准预测，辅助临床制定个体化治疗方案并提升患者生存获益 。  

Objectives of Study：

This study aims to address critical clinical challenges in the immunochemotherapy of non-small cell lung cancer (NSCLC), specifically the latency in efficacy assessment and the misdiagnosis of "pseudoprogression." We propose to construct a multi-center, multi-temporal standardized database and utilize Spatio-Temporal Graph Convolutional Networks (ST-GCN) alongside a Decomposition-Ensemble Transformer model. By deeply integrating multi-temporal CT imaging with clinical features, we aim to dynamically quantify the evolution of spatio-temporal heterogeneity within the tumor microenvironment. Ultimately, this research seeks to establish a "dynamic iRECIST" intelligent assessment standard and develop a highly interpretable automated evaluation system. These tools are designed to enable early and precise efficacy prediction, thereby assisting clinicians in formulating personalized treatment regimens and improving patient survival outcomes.

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.本研究将纳入经组织病理学或细胞学检查明确诊断为非小细胞肺癌（NSCLC）的患者，且临床分期需为III期或IV期，属于无法进行手术切除或根治性放疗的晚期病例；
2.在治疗方案上，患者必须接受以免疫检查点抑制剂（如PD-1/PD-L1抗体）为基础的联合化学治疗作为一线或后续治疗方案；
3.影像学数据的完整性是入组的核心条件，患者需拥有完整的时序CT影像资料，具体包括治疗前基线期（T0）、治疗3个疗程后（T1）以及治疗5个疗程后（T2）的CT图像，且图像层厚需＜＝1mm以满足影像组学特征提取的精度要求；
4.此外，根据RECIST 1.1标准，患者至少需存在一个可测量的靶病灶，并具备完整的临床基线资料（如年龄、性别、吸烟史、ECOG评分）及关键的分子病理数据（如EGFR/ALK基因突变状态、PD-L1表达水平等）。

Inclusion criteria

1.This study will enroll patients diagnosed with non-small cell lung cancer (NSCLC) by histopathology or cytology, with a clinical stage of III or IV, representing advanced cases that are not eligible for surgical resection or radical radiotherapy.
2.Regarding treatment regimens, patients must have received combination chemotherapy based on immune checkpoint inhibitors (PD?1/PD?L1 antibodies) as first?line or subsequent therapy.
3.The integrity of imaging data is a core inclusion criterion. Patients must have complete serial CT images, including baseline before treatment (T0), after 3 treatment cycles (T1), and after 5 treatment cycles (T2). The slice thickness of CT images must be ＜＝1 mm to meet the precision requirements for radiomic feature extraction.
4.In addition, according to RECIST 1.1 criteria, patients must have at least one measurable target lesion, complete clinical baseline data (including age, gender, smoking history, ECOG performance status), and key molecular pathological data (such as EGFR/ALK gene mutation status, PD-L1 expression level;

排除标准：

1.若患者的CT影像质量不佳，存在严重的呼吸运动伪影、心脏搏动伪影或体内金属植入物伪影，导致无法准确勾画肿瘤及瘤周区域或无法满足数据分析要求，将被排除在研究之外；
2.既往病史方面，存在其他恶性肿瘤病史的患者不予纳入；
3.临床状态上，若患者合并严重的心、肺、肝、肾功能不全或其他严重并发症，导致无法耐受免疫治疗或预期寿命过短从而影响生存期评估，亦属于排除范围；
4.为了避免对影像特征的干扰，靶病灶区域在基线影像采集前接受过放射治疗的患者也将被排除；

Exclusion criteria：

1.Patients with poor CT image quality characterized by severe respiratory motion artifacts, cardiac motion artifacts, or metallic implant artifacts that preclude accurate segmentation of the tumor and peritumoral regions or fail to meet data analysis requirements will be excluded from the study.
2.Patients with a history of other malignancies will be excluded.
3.Patients with severe cardiac, pulmonary, hepatic, or renal insufficiency, or other serious complications that result in intolerance to immunotherapy or an extremely short life expectancy (which would affect survival evaluation), will also be excluded.
4.Patients who received radiotherapy to the target lesion area before baseline imaging acquisition will also be excluded to avoid interference with imaging features.

研究实施时间：

Study execute time：

从 From 2026-03-23 00:00:00至 To 2029-02-23 00:00:00  

征募观察对象时间：

Recruiting time：

从 From 2026-03-23 00:00:00 至 To 2029-02-23 00:00:00

干预措施：

Interventions：

组别：

响应组（Responder）

样本量：

 250

Group：

Responder

Sample size：

干预措施：

非小细胞肺癌免疫联合化疗应答相关危险因素 / 影像组学特征

干预措施代码：

Intervention：

Risk factors and radiomic features associated with response to immunotherapy combined with chemotherapy in non-small cell lung cancer

Intervention code：

组别：

无响应组（Non-Responder）

样本量：

 250

Group：

Non-Responder

Sample size：

干预措施：

非小细胞肺癌免疫联合化疗非应答相关危险因素 / 影像组学特征

干预措施代码：

Intervention：

Risk factors and radiomic features associated with non-response to immunotherapy combined with chemotherapy in non-small cell lung cancer

Intervention code：

研究实施地点：

Countries of recruitment and research settings：

国家：

 中国

省(直辖市)：

 江苏省 

市(区县)：

  

Country：

China

Province：

Jiangsu

City：

单位(医院)：

 无锡市人民医院 

单位级别：

 三级甲等 

Institution
hospital：

Wuxi people’s Hospital

Level of the institution：

Tertiary A

测量指标：

Outcomes：

指标中文名：

模型预测准确性 (AUC / 灵敏度)

指标类型：

主要指标

Outcome：

Model prediction accuracy (Area Under the Curve，AUC / Sensitivity)

Type：

Primary indicator

测量时间点：

回顾性分析阶段 (模型测试时)

测量方法：

采用受试者工作特征曲线（ROC 曲线）分析法，计算模型的曲线下面积（AUC）以评估整体预测准确性；通过构建预测模型的二分类临界值，计算灵敏度（Sensitivity，真阳性率）、特异度、准确率等指标，量化模型对非小细胞肺癌免疫联合化疗应答 / 非应答的预测效能。

Measure time point of outcome：

Retrospective analysis phase (during model testing)

Measure method：

The receiver operating characteristic (ROC) curve was used to calculate the area under the curve (AUC) for evaluating the overall predictive accuracy of the model. The optimal cutoff value was determined to calculate the sensitivity (true positive rate), specificity and accuracy, so as to quantitatively evaluate the predictive efficacy of the model in identifying response or non-response to immunotherapy combined with chemotherapy in non-small cell lung cancer.

指标中文名：

无进展生存期

指标类型：

次要指标

Outcome：

Progression-Free Survival (PFS)

Type：

Secondary indicator

测量时间点：

从开始免疫联合化疗之日起，至首次出现疾病进展或任何原因死亡之日止。 随访截止日期为研究规定的终点日期。

测量方法：

根据 RECIST 1.1 标准，通过定期胸部 CT 等影像学检查评估肿瘤是否进展；采用 Kaplan?Meier 法绘制生存曲线，计算中位无进展生存期及 95% 置信区间

Measure time point of outcome：

first disease progression or death from any cause

Measure method：

Tumor progression was evaluated by regular imaging examinations (chest CT, etc.) according to RECIST 1.1 criteria. Kaplan?Meier method was used to plot survival curves, and the median progression?free survival and 95% confidence interval were calculated.

指标中文名：

客观缓解率

指标类型：

次要指标

Outcome：

Objective Response Rate ORR

Type：

Secondary indicator

测量时间点：

以末次随访时达到最佳总体疗效的时间点作为客观缓解率的判定时间点。

测量方法：

根据RECIST 1.1疗效评价标准，将治疗后肿瘤疗效分为完全缓解（CR）、部分缓解（PR）、疾病稳定（SD）、疾病进展（PD）。客观缓解率（ORR）定义为达到完全缓解（CR）和部分缓解（PR）的患者比例，通过计算符合 CR+PR 的患者数占总入组患者数的百分比进行测量。

Measure time point of outcome：

The time point at which the best overall response is achieved at the last follow-up was used as the

Measure method：

According to the RECIST 1.1 criteria, tumor response after treatment was classified into complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD).Objective response rate (ORR) was defined as the proportion of patients who achieved complete response (CR) or partial response (PR), measured by calculating the percentage of patients with CR or PR in the total enrolled patients.

指标中文名：

动态异质性指数

指标类型：

次要指标

Outcome：

Dynamic Heterogeneity Index

Type：

Secondary indicator

测量时间点：

回顾性分析阶段 (模型测试时)

测量方法：

基于治疗前与治疗后不同时间点的 CT 图像，分别勾画肿瘤靶区容积感兴趣区（VOI），提取并定量计算肿瘤内部纹理特征、密度分布及空间结构随时间变化的差异程度，综合构建动态异质性指数，以反映治疗过程中肿瘤异质性的动态变化水平。

Measure time point of outcome：

Retrospective analysis phase (during model testing)

Measure method：

Based on CT images at different time points before and after treatment, the volume of interest (VOI) of the target tumor lesion was segmented respectively. The differences in intratumoral texture features, density distribution and spatial structure over time were extracted and quantitatively calculated, and the Dynamic Heterogeneity Index (DHI) was comprehensively constructed to reflect the dynamic change of tumor heterogeneity during treatment.

采集人体标本：

Collecting sample(s)
from participants：

标本中文名：

无

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：

正在进行

Recruiting

年龄范围：

Participant age：
最小 Min age 18 岁 years
最大 Max age 岁 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）：

无

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：

无

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）：

本研究采取严格的隐私保护策略，不共享原始数据

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

This study adopts a strict privacy protection strategy and does not share raw data.

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HIS及PACS系统回顾性采集500例接受免疫化学治疗的III-IV期NSCLC患者的标准化时序CT影像与临床病理数据 。遵循“源头去标识化”与“数据不出院”原则，所有数据在录入瞬间即完成编码脱敏，并存储于物理与网络双重隔离的本地专用服务器中，密钥由PI单独离线保管 。多中心协同采用基于差分隐私与同态加密的联邦学习架构，仅传输加密模型参数而绝不共享原始数据，同时通过双盲双人背靠背标注与仲裁机制（ICC≥0.85）严格把控数据质量 。

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

Standardized sequential CT images and clinicopathological data of 500 patients with Stage III-IV NSCLC who received immunochemotherapy were retrospectively collected from the Hospital Information System (HIS) and Picture Archiving and Communication Systems (PACS). Following the principles of "de-identification at the source" and "data not leaving the hospital", all data were encoded and de-identified immediately upon entry, and stored on a local dedicated server with dual isolation of physical and network layers. The encryption key is kept separately offline by the Principal Investigator (PI). For multicenter collaboration, a federated learning framework based on differential privacy and homomorphic encryption was adopted, where only encrypted model parameters are transmitted without sharing any raw data. Meanwhile, data quality was strictly ensured through a double-blind independent double annotation and arbitration mechanism (Intraclass Correlation Coefficient [ICC] ≥ 0.85).

数据与安全监察委员会：

Data and Safety Monitoring Committee：

无/No

注册人：

Name of Registration：

  2026-03-23 17:16:13