前瞻性研究:運用基於人工智能(AI)的眼底圖像分析 以辦認眼疾並協助篩查

注册号:

Registration number:

ChiCTR2400080231 

最近更新日期:

Date of Last Refreshed on:

2024-12-10 15:07:39 

注册时间:

Date of Registration:

2024-01-24 00:00:00 

注册号状态:

补注册

Registration Status:

Retrospective registration

注册题目:

前瞻性研究:運用基於人工智能(AI)的眼底圖像分析 以辦認眼疾並協助篩查

Public title:

Using an artificial intelligence based ocular image analysis for eye disease identification to support eye disease screening: a prospective study

注册题目简写:

English Acronym:

研究课题的正式科学名称:

前瞻性研究:運用基於人工智能(AI)的眼底圖像分析 以辦認眼疾並協助篩查

Scientific title:

Using an artificial intelligence based ocular image analysis for eye disease identification to support eye disease screening: a prospective 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:

蔡珊珊女士 

研究负责人:

張艷蕾副教授 / 譚智勇教授 

Applicant:

Ms Jennifer Tsoi  

Study leader:

Dr Cheung Yim Lui Carol / Prof Tham Chee Yung Clement 

申请注册联系人电话:

Applicant telephone:

+852 3943 5818

研究负责人电话:

Study leader's
telephone:

+852 3943 5831

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:

jennifertsoi@cuhk.edu.hk

研究负责人电子邮件:

Study leader's E-mail:

carolcheung@cuhk.edu.hk

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:

九龍亞皆老街147K號 香港眼科醫院3樓

研究负责人通讯地址:

九龍亞皆老街147K號香港眼科醫院4樓

Applicant address:

3/F, Hong Kong Eye Hospital, 147K Argyle Street, Kowloon

Study leader's address:

4/F, Hong Kong Eye Hospital, 147K Argyle Street, Kowloon

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:

香港中文大學眼科及視覺科學學系

Applicant's institution:

Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong

研究负责人所在单位:

香港中文大學眼科及視覺科學學系

Affiliation of the Leader:

Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong

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:

KC/KE-20-0241/ER-1

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:

九龍中及九龍東聯網臨床研究倫理委員會

Name of the ethic committee:

Research Ethics Committee (Kowloon Central / Kowloon East)

伦理委员会批准日期:

Date of approved by ethic committee:

2020-11-26 00:00:00

伦理委员会联系人:

Ms Lyon Chan

Contact Name of the ethic committee:

Ms Lyon Chan

伦理委员会联系地址:

香港九龍加士居道30號伊利沙伯醫院護士宿舍4樓414室

Contact Address of the ethic committee:

Room 414, Nurse Quarters, Queen Elizabeth Hospital, 30 Gascoigne Road, Kowloon, Hong Kong

伦理委员会联系人电话:

Contact phone of the ethic committee:

+852 3506 8888

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:

香港中文大學眼科及視覺科學學系

Primary sponsor:

Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong

研究实施负责(组长)单位地址:

香港九龍亞皆老街147K號香港眼科醫院3樓

Primary sponsor's address:

3/F, Hong Kong Eye Hospital, 147K Argyle Street, Kowloon, Hong Kong

试验主办单位(项目批准或申办者):

Secondary sponsor:

国家:

香港

省(直辖市):

香港

市(区县):

Country:

Hong Kong

Province:

Hong Kong

City:

单位(医院):

香港中文大學眼科及視覺科學學系

具体地址:

香港九龍亞皆老街147K號香港眼科醫院3樓

Institution
hospital:

Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong

Address:

3/F, Hong Kong Eye Hospital, 147K Argyle Street, Kowloon, Hong Kong

经费或物资来源:

部門經費

Source(s) of funding:

Departmental fund

研究疾病:

與視網膜和視神經頭相關的眼疾(例如糖尿病黃斑水腫、青光眼和其他視網膜異常)  

Target disease:

Eye diseases related to retina and optic nerve head (e.g., DME, glaucoma, and other retinal abnormalities)

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:

诊断试验新技术临床试验 

Study phase:

Diagnostic New Technique Clincal Study

研究设计:

诊断试验诊断准确性 

Study design:

Diagnostic test for accuracy 

研究目的:

本研究旨在為評估AI輔助系統於自動評估OCT影像質素的表現、從OCT影像辨認出與視網膜及視神經相關的眼疾(例如糖尿病黃斑水腫、青光眼和其他視網膜疾病),並提供臨床分流和轉診建議,以協助現實中對眼睛疾病之篩查。  

Objectives of Study:

The current project aims to assess the performance of the AI-assisted system prospectively for automatically assessing image quality, identifying eye diseases related to retina and optic nerve (e.g. DME, glaucoma, and other retinal abnormalities) from OCT images, and providing a clinical triage and referral suggestion to support real-world eye disease screening.

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:

接受OCT檢查有困難的參與者

Exclusion criteria:

Subjects who have difficulty receiving OCT examinations

研究实施时间:

Study execute time:

From 2020-11-01 00:00:00 To 2026-10-31 00:00:00  

征募观察对象时间:

Recruiting time:

From 2021-02-01 00:00:00 To 2025-10-31 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):

- 使用二維非立體數位眼底攝影進行糖尿病眼疾篩檢 - 青光眼篩檢和 AMD 篩檢主要透過眼科診所或定期眼科檢查的人群進行機會性病例識別來進行 - 由專業評級師或眼科醫生人手審查 OCT 影像

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):

- Screening programmes for diabetic eye disease using 2-dimensional non-stereoscopic digital fundus photography - Glaucoma screening and AMD screening are largely conducted through opportunistic case identification in eye clinics, or amongst people having regular eye check-ups - Manual review of OCT images by professional graders or ophthalmologists

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):

前瞻性地評估人工智能輔助系統的性能,自動評估影像質量,從 OCT 影像中識別與視網膜和視神經頭相關的眼部疾病(例如 DME、青光眼和其他視網膜異常),並提供臨床分流和轉診建議以支援現實世界的眼部病篩檢。AI輔助系統將首先評估OCT影像質素,過濾不可評級的OCT影像,並允許可評級的OCT影像作進一步分析。 然後,AI系統將評估3D體積OCT數據,以檢測和識別OCT影像中是否存在任何眼部疾病。AI輔助系統最終會對識別出的眼部疾病提供臨床分流建議(例如緊急、次緊急、非緊急),以便於三級眼科診所及時治療和管理。

Index test:

The current project aims to assess the performance of the AI-assisted system prospectively for automatically assessing image quality, identifying eye diseases related to retina and optic nerve head (e.g., DME, glaucoma, and other retinal abnormalities) from OCT images, and providing a clinical triage and referral suggestion to support real-world eye disease screening. The AI-assisted system will first assess OCT image quality to filter out ungradable OCT images, and allow gradable OCT images for further analysis. The AI system will then assess the 3D volumetric OCT data to detect and identify whether there are any eye diseases present in the OCT images. The AI-assisted system will finally provide a clinical triage suggestion (e.g., urgent referral, semi-urgent referral, non-urgent referral) for the identified eye diseases to facilitate timely treatment and management at tertiary eye clinics.

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)

1) 抱怨明顯的視力障礙或視力測試中自我報告的變化(例如阿姆斯勒方格表); 2) 從社區篩檢計畫、初級保健機構、地區健康中心或非政府組織獲知異常檢測結果; 3) 有青光眼家族史;或有青光眼相關的可疑發現(例如視杯與視盤比率增加、眼壓升高、視盤出血和視網膜神經纖維層變薄); 4) 有糖尿病家族史或確診患有糖尿病;或患有任何糖尿病性眼疾(例如糖尿病黃斑水腫、糖尿病性視網膜病變); 5) 患有任何視網膜異常(例如,老年黃斑部病變、視網膜前膜、黃斑裂孔)。

例数:

Sample size:

2000

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):

1) With complaints about perceptible visual impairment or self-reported alternations in their vision test (e.g., Amsler grid); 2) Informed with abnormal test results from community-based screening programs, primary care settings, DHCs or NGOs; 3) With a family history of glaucoma; or with glaucoma-related suspicious findings (e.g. increased cup to disc ratio, increased intraocular pressure, disc hemorrhage, and retinal nerve fibre layer thinning); 4) With a family history or diagnosed with diabetes mellites; or with any diabetic eye diseases (e.g. diabetic macular edema, diabetic retinopathy); 5) With any retinal abnormalities (e.g., age-related macular degeneration, epiretinal membrane, macular hole).

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:

N/A

研究实施地点:

Countries of recruitment and research settings:

国家:

香港

省(直辖市):

香港 

市(区县):

 

Country:

Hong Kong

Province:

Hong Kong

City:

单位(医院):

香港眼科醫院 

单位级别:

無 

Institution
hospital:

Hong Kong Eye Hospital

Level of the institution:

N/A

国家:

香港

省(直辖市):

香港 

市(区县):

 

Country:

Hong Kong

Province:

Hong Kong

City:

单位(医院):

香港中文大學眼科中心 

单位级别:

無 

Institution
hospital:

CUHK Eye Centre

Level of the institution:

N/A

测量指标:

Outcomes:

指标中文名:

人工智能輔助系統在評估影像品質、檢測眼疾和提供分流/轉診建議方面的準確性。

指标类型:

主要指标

Outcome:

The accuracy of the AI-assisted system for image quality assessment, eye disease detection and triage/referral suggestion.

Type:

Primary indicator

测量时间点:

一次訪問

测量方法:

基於AI輔助系統的OCT影像分析

Measure time point of outcome:

1 visit

Measure method:

OCT image analysis based on the AI-assisted system.

指标中文名:

人工智能輔助系統升級後在影像品質評估和眼部疾病檢測的準確性

指标类型:

次要指标

Outcome:

The accuracy of the upgraded AI-assisted system for image quality assessment and eye disease detection

Type:

Secondary 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:

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:

正在进行

Recruiting

年龄范围:

Participant age:

最小 Min age 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:

公开/Public

盲法:

Blinding:

试验完成后的统计结果(上传文件):

Calculated Results after
the Study Completed(upload file):

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):

研究發表後半年上傳至ResMan(www.medresman.org.cn)

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

Upload the research to ResMan (www.medrescman. org. cn) half a year after publication

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:

OCT 影像:使用市售 OCT 設備(例如 Cirrus Carl Zeiss Meditec, Inc, Dublin, CA;DRI-OCT, Topcon, Inc, Tokyo, Japan;Spectralis, Heidelberg Engineering, Heidelbery, German)進行視盤和黃斑立方體掃描對所有招募的受試者的雙眼進行檢查。 AI輔助系統:捕捉眼底和OCT影像後,眼科掃描後立即從眼部影像資料庫擷取原始影像資料到特定位置。然後,資料將自動傳輸到 GPU 伺服器,並由人工智慧輔助系統進一步處理。人工智慧輔助系統由四個部分組成。第一部分用於評估影像品質並將 FP 影像和 OCT 掃描分類為「可分級或不可分級」。第二部分用於將可分級眼底影像分類為 1)「是/否可參考的糖尿病視網膜病變」和 2)「是/否威脅視力的糖尿病視網膜病變;可分級黃斑 OCT 掃描 1)“是/否 DME”,以及 2)“是/否視網膜異常”;可分級視盤 OCT 掃描 1)「是/否青光眼性視神經病變」和 2)「是/否近視特徵」。 AI輔助系統也會根據影像品質的分類和眼部疾病的識別,輸出轉診/分診建議(緊急、半緊急或非緊急)。 招募時的其他資料收集:針孔視力、未使用藥物擴瞳的便攜式或桌上型眼底機(例如Topcon NW500、TRC 50DX、Topcon Inc)的FP、基線人口統計資料(例如種族、性別、吸煙、體重、身高)、系統性也將獲得病史(例如高血壓、糖尿病、高血脂)、認知篩檢測試[例如蒙特利爾認知評估5 分鐘(MoCA 5 分鐘) 方案和修改後的確定癡呆8 (mAD8 ) 問卷]。 電子數據將只在我們安全的研究室電腦內保存,並受到密碼保護。這項研究的資料將給予香港中文大學醫學院眼科及視覺科學學系進行統計分析。參與者的身份將受嚴格保密,只有整體的結果將被公佈。個人資料將於研究完結後保存五年。於任何時間,參與者可要求銷毀所有相關的研究結果和記錄。如無需用於其他研究或治療,個人資料將於研究完成後保存5年。 我們將極為謹慎地處理患者數據,避免以任何形式侵犯患者的隱私。 為保障病人私隱,所有研究資料會依照醫院管理局/醫院處理/儲存/銷毀病人病歷的政策處理。 USB 裝置不會用於儲存病患資訊或個人資料。 個人資料(姓名、香港身分證、OPD/醫院號碼、地址及任何其他個人識別資料)不會記錄在項目的資料表或電子文件中。 將使用研究代碼來代替。 包含研究代碼和患者身份之間的連結資訊的電子文件文件不會包含任何其他信息,並且將與研究數據文件或數據表分開保存,具有與醫療記錄相同的嚴格安全性。任何包含個人識別資訊的文件或電子文件都將被視為醫療記錄的一部分,並將根據醫院政策遵守同樣嚴格的安全規定。所有調查人員將負責資料處理和保護。所有可辨識的個人資料均會匿名化,並遵守醫管局處理病人資料私隱的政策。

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

OCT imaging: Optic disc and macular cube scans from commercially available OCT devices (e.g. Cirrus Carl Zeiss Meditec, Inc, Dublin, CA; DRI-OCT, Topcon, Inc, Tokyo, Japan; Spectralis, Heidelberg Engineering, Heidelbery, Germany) will be performed on both eyes for all recruited subjects. AI-assisted system: After capturing the fundus and OCT images, raw image data will be extracted from the ocular imaging database to a specific location immediately after ophthalmic scanning. The data will then be automatically transfer to a GPU server for further processing by the AI-assisted system. The AI-assisted system is comprised of four parts. The first part is for assessing image quality and classifying FP images and OCT scans into "gradable or ungradable". The second part is for classifying gradable fundus images into 1) "yes/no referable diabetic retinopathy", and 2) "yes/no vision-threatening diabetic retinopathy; gradable macula OCT scans into 1) "yes/no DME", and 2) "yes/no retinal abnormalities"; gradable optic disc OCT scans into 1)"yes/no glaucomatous optic neuropathy", and 2) "yes/no myopic features". A referral/triage suggestion (urgent, semi-urgent, or non-urgent) will also be outputted by the AI-assisted system according to the classifications of image quality and identification of eye disease. Other data collection at recruitment: Pinhole visual acuity, FPs from portable or desktop fundus machines without pharmaceutical pupil dilation (e.g., Topcon NW500, TRC 50DX, Topcon Inc), baseline demographics (e.g. ethnicity, gender, smoking, weight, height), systemic medical history (e.g. hypertension, diabetes, hyperlipidemia), cognitive screening test [e.g., Montreal Cognitive Assessment 5-minute (MoCA 5-min) protocol and modified Ascertain Dementia 8 (mAD8) questionnaire] will also be obtained. Electronic data will be only saved in physically-secured and password-protected computers in the research office of CUHK Eye Centre. Information from this study will be submitted to the Department of Ophthalmology & Visual Sciences, the Chinese University of Hong Kong for statistical analysis. Only the overall result will be published and subjects' identity will remain confidential. The signed consent form will be stored separately from interview notes and personal data to further protect confidentiality. Access to the data will be restricted to the researchers of this study. Records and results of all study investigations can be destroyed on subjects' request in future. Personal data will be kept for 5 years after study completion if not for further research or clinical use. Patient data would be handled with utmost care taking care not to breach patient's privacy in any form. To protect patient privacy, all research data would be handled in line with Hospital Authority / Hospital’s policy in handling / storage / destruction of patients’ medical records. USB Device would not be used for patient information nor personal data. Personal data (name, HKID, OPD / hospital numbers, address and any other personal identifiable information) would not be recorded on the project’s data sheets or electronic files. A study code would be used instead. The document of electronic file containing the linkage information between the study code and the identity of the patient would not contain any other information and would be kept separate from the study data files or data sheets with the same stringent security as the medical record. Any documents or electronic files containing personal identifiable information would be considered as part of the medical record and would be dealt with the same stringent regulations of security according to the hospital policies. All the investigators would be responsible for data handling and protection. All identifiable personal data will be anonymised and will follow the HA policy on handling of patient data privacy.

数据与安全监察委员会:

Data and Safety Monitoring Committee:

暂未确定/Not yet

注册人:

Name of Registration:

 2024-01-24 14:28:21