中国临床试验注册中心 - 世界卫生组织国际临床试验注册平台一级注册机构

注册号： Registration number：	ChiCTR2600122168
最近更新日期： Date of Last Refreshed on：	2026-04-09 16:39:29
注册时间： Date of Registration：	2026-04-09 00:00:00
注册号状态：	预注册
Registration Status：	Prospective registration
注册题目：	基于时空特征融合的抑郁症消极风险语音预测模型研究
Public title：	Research on Depression Negative Risk Speech Prediction Model Based on Spatio-Temporal Feature Fusion
注册题目简写：
English Acronym：
研究课题的正式科学名称：	基于时空特征融合的抑郁症消极风险语音预测模型研究
Scientific title：	Research on Depression Negative Risk Speech Prediction Model Based on Spatio-Temporal Feature Fusion
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：	高莉玲	研究负责人：	高莉玲
Applicant：	Liling Gao	Study leader：	Liling Gao
申请注册联系人电话： Applicant telephone：	+86 13865912788	研究负责人电话： Study leader's telephone：	+86 551 6361 6012
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：	lilinggao868@sina.com	研究负责人电子邮件： Study leader's E-mail：	lilinggao868@sina.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：	安徽省合肥市蜀山区黄山路316号	研究负责人通讯地址：	安徽省合肥市蜀山区黄山路316号
Applicant address：	316 Huangshan Road, Shushan District, Hefei City, Anhui Province	Study leader's address：	316 Huangshan Road, Shushan District, Hefei City, Anhui Province
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：	合肥市第四人民医院
Applicant's institution：	Hefei Fourth People's Hospital
研究负责人所在单位：	合肥市第四人民医院
Affiliation of the Leader：	Hefei Fourth Peoples 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：	HFSY-IRB-YJ-KYXM-GLL（2026-020-001）	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：	合肥市第四人民医院医学伦理委员会
Name of the ethic committee：	Ethics Review Committee of Hefei Fourth People's Hospital
伦理委员会批准日期： Date of approved by ethic committee：	2026-02-05 00:00:00
伦理委员会联系人：	曹娜娜
Contact Name of the ethic committee：	Cao NaNa
伦理委员会联系地址：	安徽省合肥市蜀山区黄山路316号
Contact Address of the ethic committee：	316 Huangshan Road, Shushan District, Hefei City, Anhui Province
伦理委员会联系人电话： Contact phone of the ethic committee：	+86 551 63616193	伦理委员会联系人邮箱： Contact email of the ethic committee：	2514331322@qq.com

研究实施负责（组长）单位：

合肥市第四人民医院

Primary sponsor：

Hefei Fourth Peoples Hospital

研究实施负责（组长）单位地址：

安徽省合肥市蜀山区黄山路316号

Primary sponsor's address：

316 Huangshan Road, Shushan District, Hefei City, Anhui Province

试验主办单位(项目批准或申办者)：

Secondary sponsor：

国家：	中国	省(直辖市)：	安徽省	市(区县)：
Country：	China	Province：	Anhui	City：
单位(医院)：	合肥市第四人民医院	具体地址：	安徽省合肥市蜀山区黄山路316号
Institution hospital：	Hefei Fourth Peoples Hospital	Address：	316 Huangshan Road, Shushan District, Hefei City, Anhui Province

经费或物资来源：

2025年安徽省合肥市卫生健康科技项目

Source(s) of funding：

Hefei City Health Technology Project of Anhui Province, Year 2025

研究疾病：

抑郁症

Target disease：

Major Depressive Disorder (MDD)

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：

其它

Study phase：

N/A

研究设计：

病例对照研究

Study design：

Case-Control study

研究目的：

本研究旨在针对当前抑郁症消极风险评估依赖主观问诊与量表、缺乏客观高效工具的临床困境，基于语音信号分析与深度学习技术，构建一种基于跨位置聚合网络（CPANet）与时间卷积网络（TCN）融合的抑郁症消极风险语音预测模型。通过挖掘与消极风险（如自杀意念）相关的关键声学特征，实现对患者消极风险水平的客观、精准识别与早期预警，从而为临床提供一种非侵入性、易推广的辅助筛查工具，提升抑郁症风险识别的及时性与准确性，降低漏诊率和自杀风险。

Objectives of Study：

This study aims to address the current clinical challenge in depression negative risk assessment, which relies on subjective interviews and scales and lacks objective and efficient tools. Based on speech signal analysis and deep learning technology, a speech-based prediction model for depression negative risk is constructed by integrating a Cross-location Aggregation Network (CPANet) and a Temporal Convolutional Network (TCN). By mining key acoustic features associated with negative risk (such as suicidal ideation), the study seeks to achieve objective and accurate identification and early warning of patients' negative risk levels. This will provide a non-invasive, easy-to-promote auxiliary screening tool for clinical practice, enhance the timeliness and accuracy of depression risk identification, and reduce missed diagnosis rates and suicide risk.

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

1. Major Depression High Suicide Risk Group:
(1) Meets DSM-5 diagnostic criteria for depression and is independently diagnosed by two psychiatrists of associate senior level or above, reaching a consensus.
(2) Assessed as high risk using the Suicide Ideation Scale (SSI), defined as an SSI total score ≥6 or the presence of clear suicidal thoughts, plans, or behaviors.
(3) Aged 18-60, of Han ethnicity, any gender.
(4) At least junior high school education, able to understand the study content and communicate effectively.
(5) Voluntarily participate and sign written informed consent.

2. Major Depression Low Suicide Risk Group:
Inclusion criteria are basically the same as the high-risk group, with the core difference being risk assessment: SSI total score <6, and no clear suicidal thoughts, plans, or behaviors.

3. Healthy Control Group:
(1) Matched with the case groups in terms of age, gender, and education level.
(2) At least junior high school education, normal audiovisual ability.
(3) Confirmed to have no current or past history of any mental disorder through standardized mental examinations and diagnostic interviews.
(4) Voluntarily participate and sign written informed consent.

排除标准：

1. 病例组排除标准： (1)患有严重未控制的躯体疾病（如心血管、呼吸、神经系统器质性疾病），可能影响语音或认知功能。 (2)有精神活性物质依赖或滥用史，或入组前三个月内有使用行为。 (3)存在严重冲动行为或当前处于自杀高危状态需紧急干预者。 (4)入组前一个月内接受过改良电抽搐治疗（MECT）。 (5)正在服用可能显著影响语言功能的药物（如大剂量镇静药、抗胆碱能药）。 (6)合并其他精神障碍诊断（如精神分裂症、双相障碍等）。 (7)存在严重的言语或听力障碍。 2. 健康对照组排除标准： (1)有任何精神障碍的当前或既往史，或有一级/二级亲属的精神障碍家族史。 (2)存在听力、言语表达障碍或发音器官疾病。 (3)当前正在服用影响精神状态或语音表达的药物。 (4)近三个月内经历重大生活应激或创伤性事件。

Exclusion criteria：

1. Exclusion criteria for the patient group: (1) Suffering from severe uncontrolled physical diseases (such as cardiovascular, respiratory, or organic neurological diseases) that may affect speech or cognitive function. (2) History of dependence on or abuse of psychoactive substances, or any use within three months prior to enrollment. (3) Presence of severe impulsive behavior or currently in a high-risk suicidal state requiring urgent intervention. (4) Received modified electroconvulsive therapy (MECT) within one month before enrollment. (5) Currently taking medications that may significantly affect language function (such as high-dose sedatives or anticholinergic drugs). (6) Comorbid other psychiatric disorder diagnoses (such as schizophrenia, bipolar disorder, etc.). (7) Presence of severe speech or hearing impairment. 2. Exclusion criteria for the healthy control group: (1) Current or past history of any psychiatric disorder, or a family history of psychiatric disorders in first- or second-degree relatives. (2) Presence of hearing, speech expression disorders, or diseases of the articulatory organs. (3) Currently taking medications that affect mental state or speech expression. (4) Experienced major life stressors or traumatic events within the past three months.

研究实施时间：

Study execute time：

从 From 2026-04-30 00:00:00至 To 2027-12-31 00:00:00

征募观察对象时间：

Recruiting time：

从 From 2026-04-30 00:00:00 至 To 2027-12-31 00:00:00

干预措施：

Interventions：

组别：	抑郁症低自杀风险组	样本量：	100
Group：	Low Suicide Risk Depression Group	Sample size：	100
干预措施：	无	干预措施代码：
Intervention：	no	Intervention code：

组别：	抑郁症高自杀风险组	样本量：	100
Group：	High Suicide Risk Depression Group	Sample size：	100
干预措施：	无	干预措施代码：
Intervention：	no	Intervention code：

组别：	健康对照组	样本量：	100
Group：	Healthy Control Group	Sample size：	100
干预措施：	无	干预措施代码：
Intervention：	no	Intervention code：

研究实施地点：

Countries of recruitment and research settings：

国家：	中国	省(直辖市)：	安徽省	市(区县)：
Country：	China	Province：	Anhui	City：
单位(医院)：	合肥市第四人民医院	单位级别：	三级甲等
Institution hospital：	Hefei Fourth Peoples Hospital	Level of the institution：	Tertiary A

测量指标：

Outcomes：

指标中文名：	临床表型	指标类型：	次要指标
Outcome：	clinical phenotype	Type：	Secondary indicator
测量时间点：	2026.5	测量方法：	量表及语音数据分析
Measure time point of outcome：	2026.5	Measure method：	Scales and speech data analysis

指标中文名：	诊断与严重程度	指标类型：	主要指标
Outcome：	Diagnosis and Severity	Type：	Primary 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：	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：

尚未开始

Not yet recruiting

年龄范围：

Participant age：

最小 Min age	18	岁 years
最大 Max age	60	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）：	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：	本研究的数据采集与管理遵循标准化操作规程，以确保数据的完整性、准确性和安全性。数据采集涵盖临床评估数据与语音数据两部分：临床评估通过结构化访谈及标准化量表（如汉密尔顿抑郁量表、自杀意念量表等）收集患者的抑郁症状严重程度、自杀风险水平及人口学信息；语音数据则在隔音录音室内使用高保真设备采集，包括结构化访谈和自然对话两种模式，采样率不低于16 kHz。所有数据均由经过统一培训的研究人员进行采集与录入，并采用双人核对机制确保数据准确性。数据管理方面，建立加密数据库对临床资料与语音文件进行统一存储与管理，去除个人身份信息后以唯一编码标识，保障受试者隐私。语音数据经过预处理（如降噪、端点检测）后，提取多维声学特征并建立特征库，为后续建模分析提供高质量数据基础。整个数据采集与管理过程严格遵守医学研究伦理要求，并接受伦理委员会的监督。
Data collection and Management (A standard data collection and management system include a CRF and an electronic data capture：	This study adheres to standardized operating procedures for data collection and management to ensure data integrity, accuracy, and security. Data collection encompasses both clinical assessment data and speech data: clinical assessments gather information on the severity of depressive symptoms, suicide risk levels, and demographic details through structured interviews and standardized scales (e.g., the Hamilton Depression Rating Scale, Scale for Suicidal Ideation). Speech data are collected in a soundproof recording room using high-fidelity equipment, employing two modes—structured interviews and natural conversations—with a sampling rate no lower than 16 kHz. All data are collected and entered by uniformly trained researchers, with a double-check mechanism to ensure accuracy. For data management, an encrypted database is established for the unified storage and management of clinical records and speech files. Personal identifiers are removed and replaced with unique codes to protect participant privacy. After preprocessing (e.g., noise reduction, voice activity detection), multi-dimensional acoustic features are extracted from the speech data to build a feature library, providing a high-quality data foundation for subsequent modeling and analysis. The entire data collection and management process strictly complies with medical research ethics requirements and is subject to oversight by the ethics committee.
数据与安全监察委员会： Data and Safety Monitoring Committee：	有/Yes
注册人： Name of Registration：	2026-04-09 16:39:20