Records View

Diagnosis of lung nodule and lung cancer on screening chest radiographs: Comparative clinical trial for evaluation of artificial intelligence-integrated PACS versus conventional PACS

Status Approved

  • First Submitted Date

    2020/05/21

  • Registered Date

    2020/05/22

  • Last Updated Date

    2020/11/24

CRIS Required

WHO ICTRP (International Clinical Trial Registry Platform) Required

  • 1. Background

    Background - CRIS Registration Number, Unique Protocol ID, Public/Brief Title, Scientific Title, Acronym, MFDS Regulated Study, IND/IDE Protocol, Registered at Other Registry, Name of Registry/Registration Number
    CRIS
    Registration Number
    KCT0005051
    Unique Protocol ID D-1908-160-1059
    Public/Brief Title Diagnosis of lung nodule and lung cancer on screening chest radiographs: Comparative clinical trial for evaluation of artificial intelligence-integrated PACS versus conventional PACS
    Scientific Title Diagnosis of lung nodule and lung cancer on screening chest radiographs: Comparative clinical trial for evaluation of artificial intelligence-integrated PACS versus conventional PACS
    Acronym
    MFDS Regulated Study No
    IND/IDE Protocol No
    Registered at Other Registry No
    Healthcare Benefit Approval Status Not applicable
  • 2. Institutional Review Board / Ethics Committee

    Institutional Review Board Information
    Board Approval Status Submitted approval
    Board Approval Number D-1908-160-1059
    Approval Date 2020-05-13
    Institutional Review Board Name Seoul National University Hospital Institutional Review Board
    Institutional Review Board Address 103, Daehak-ro, Jongno-gu, Seoul
    Institutional Review Board Telephone 02-2072-0694
    Data Monitoring Committee No
  • 3. Contact Details

    Contact Details Information - Contact Person for Principal Investigator / Scientific Queries, Contact Person for Public Queries, Contact Person for Updating Information의 Name, Title, Email, Telephone, Cellular Phone, Affiliation, Address
    Contact Person for Principal Investigator / Scientific Queries
    Name Jin Mo Goo
    Title Professor
    Telephone +82-2-2072-2624
    Affiliation Seoul National University Hospital
    Address 101 Daehak-ro, Jongno-gu, Seoul
    Contact Person for Public Queries
    Name Ju Gang Nam
    Title Clinical Professor
    Telephone +82-2-2072-2254
    Affiliation Seoul National University Hospital
    Address 101 Daehak-ro, Jongno-gu, Seoul
    Contact Person for Updating Information
    Name Ju Gang Nam
    Title Clinical Professor
    Telephone +82-2-2072-2254
    Affiliation Seoul National University Hospital
    Address 101 Daehak-ro, Jongno-gu, Seoul
  • 4. Status

    Status Information - Study Site, Overall Recruitment Status, Date of First Enrollment, Status of First Enrollment, Target Number of Participant, Primary Completion Date, Recruitment Status by Participating Study Site, Name of Study Site, Recruitment Status, Date of First Enrollment, Status of First Enrollemnt
    Study Site Single
    Overall Recruitment Status Recruiting
    Date of First Enrollment 2020-06-15 Actual
    Target Number of Participant 84000
    Primary Completion Date 2021-09-30 , Anticipated
    Study Completion Date 2021-12-31 , Anticipated
    Recruitment Status by Participating Study Site 1
    Name of Study Seoul National University Hospital
    Recruitment Status Recruiting
    Date of First Enrollment 2020-06-15 ,
  • 5. Source of Monetary / Material Support

    Source of Monetary / Material Support Information - Organization Name, Organization Type, Project ID
    1. Source of Monetary/Material Support
    Organization Name Ministry of Health & Welfare
    Organization Type Government
    Project ID HI19C1129
  • 6. Sponsor Organization

    Sponsor Organization Information - Organization Name, Organization Type
    1. Sponsor Organization
    Organization Name Seoul National University Hospital
    Organization Type Medical Institute
  • 7. Study Summary

    Study Summary Information
    Lay Summary
    Background: Lung cancer is the highest mortality-cancer in the world, and it is well known that lung cancer screening is important as early detection yields improved survival. While large randomized controlled trials have demonstrated the efficacy of screening with low-dose CT in high-risk groups, chest radiograph has failed to show effectiveness as a screening modality, and it has been reported that about 30% of lung cancers were missed in chest radiography on formal reading. However, chest radiograph is still the most commonly performed imaging modality, and has many advantages in terms of accessibility, cost, and radiation exposure. There are many causes of miss-interpretation of lung cancer in chest radiograph, but the location of the lesion and the fatigue of the radiologists are well known causes. Recently, many retrospective studies have reported that artificial intelligence-algorithms using deep learning may compensate these factors. However, prospective clinical trials have not yet been conducted to show the improvement of the interpretation-quality when using deep learning algorithms compared to the radiologists reading alone in actual clinical situations. In order to apply the algorithm in actual workflow, first there must be a quality-approved algorithm, and second the PACS system should be integrated with the algorithm. In this study, we aim to evaluate if the artificial intelligence-integrated PACS (Infinitt M6) system integrated with the lung lesion-detection algorithm (Lunit INSIGHT CXR MCA), which has been verified through various studies and approved by the Ministry of Food and Drug Safety, may improve lung nodule detection rate from chest radiographs in an asymptomatic health checkup population.
    
    Hypothesis: Artificial intelligence-integrated PACS would increase the lung nodule detection rate from chest radiographs on an asymptomatic health checkup population.
    
    Objective: To investigate the effectiveness of the Artificial intelligence-integrated PACS in lung nodule detection from chest radiographs on an asymptomatic health checkup population, and help set up clinical workflow.
    
    Plan: All chest radiographs from asymptomatic patients who visit Seoul National University healthcare screening center are randomly distributed to conventional PACS system and Artificial intelligence-integrated PACS system for interpretation. All other reading conditions (reading environment, reading doctor, interpretation guideline) should be equal. The lung nodule detection rate, i.e. rate of positive cases who are confirmed to have meaningful nodules on the follow-up chest CT (taken within 3 months) divided by total population, is evaluated and compared between two groups. Additionally, false positive rate, true positive rate, false negative rate, lung cancer detection rate, and other lung disease detection rate are evaluated and compared between two groups for available cases.
  • 8. Study Design

    Study Design Information - Study Type, Study Purpose, Phase, Intervention Model, Blinding/Masking, Blinded Subject, Allocation, Intervention Type, Intervention Description, Number of Arms, Arm Label, Target Number of Participant, Arm Type, Arm Description
    Study Type Interventional Study
    Study Purpose
    Diagnosis
    Phase Not applicable
    Intervention Model Parallel  
    Blinding/Masking Open
    Allocation RCT
    Intervention Type Medical Device  
    Intervention Description
    In case group, chest radiograph interpretation will be performed using a deep learning-based computer-aided detection software (Software name: Lunit INSIGHT CXR MCA, Manufacturer: Lunit Inc.). The software was approved as a medical device by the Ministry of Food and Drug Safety, Republic of Korea. The software was designed to analyze a chest radiograph image, and identify relevant abnormalities (nodule, consolidation, and pneumothorax) on the image, and to assist the interpretation by doctors. The software provides a probability value for the presence of abnormality on the chest radiograph, and localization of abnormality overlaid on the original image. The duty board-certified radiologist will interpret chest radiographs after evaluation of both original chest radiograph image and the result from the computer-aided detection system.
    In control group, the duty board-certified radiologist will interpret chest radiographs after evaluation of original chest radiograph images only, as conventional clinical practice.
    Number of Arms 2
    Arm 1

    Arm Label

    Artificial intelligence computer-aided detection system-assisted interpretation group

    Target Number of Participant

    42000

    Arm Type

    Experimental

    Arm Description

    Chest radiograph interpretation will be performed using a deep learning-based computer-aided detection software (Software name: Lunit INSIGHT CXR MCA, Manufacturer: Lunit Inc.). The software was approved as a medical device by the Ministry of Food and Drug Safety, Republic of Korea. The software was designed to analyze a chest radiograph image, and identify relevant abnormalities (nodule, consolidation, and pneumothorax) on the image, and to assist the interpretation by doctors. The software provides a probability value for the presence of abnormality on the chest radiograph, and localization of abnormality overlaid on the original image. The board-certified radiologist will interpret chest radiographs after evaluation of both original chest radiograph image and the result from the computer-aided detection system.
    Arm 2

    Arm Label

    Conventional interpretation group

    Target Number of Participant

    42000

    Arm Type

    Active comparator

    Arm Description

    The duty board-certified radiologist will interpret chest radiographs after evaluation of original chest radiograph images only, as conventional clinical practice.
  • 9. Subject Eligibility

    Subject Eligibility Information
    Condition(s)/Problem(s) * (J00-J99)Diseases of the respiratory system 
       (J98.40)Solitary pulmonary nodule 

    Lung nodule (non-calcified solid nodules larger than or equal to 8 mm or part solid-nodules with solid portion larger than or equal to 6 mm)
    Rare Disease No
    Inclusion Criteria

    Gender

    Both

    Age

    19Year~No Limit

    Description

    All adult individuals who performed chest radiograph in Seoul National University Healthcare Screening Center
    Exclusion Criteria
    Individuals who had respiratory symptoms when taking chest radiograph
    Healthy Volunteers Yes
  • 10. Outcome Measure(s)

    Outcome Measure(s) Information - Type of Primary Outcome, Primary Outcome, Outcome, Timepoint, Secondary Outcome, Outcome, Timepoint
    Type of Primary Outcome Efficacy
    Primary Outcome(s) 1
    Outcome
    Lung nodule detection rate
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 1
    Outcome
    Positive rate
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 2
    Outcome
    Sensitivity
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 3
    Outcome
    Specificity
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 4
    Outcome
    Positive predictive value
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 5
    Outcome
    Negative predictive value
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 6
    Outcome
    Lung cancer detection rate
    Timepoint
    End of the trial
    Secondary Outcome(s) 7
    Outcome
    Detection rate of other lung disease
    Timepoint
    End of the trial
    Secondary Outcome(s) 8
    Outcome
    False referral rate
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 9
    Outcome
    Revisit rate within three months
    Timepoint
    Three months after chest radiograph
    Secondary Outcome(s) 10
    Outcome
    CT referral rate
    Timepoint
    Three months after chest radiograph
  • 11. Study Results and Publication

    Study Results and Publication Information - Result Registered, Final Enrollment Number, Number of Publication, Publications, Results Upload, Date of Posting Results, Protocol URL or File Upload, Brief Summary
    Result Registered No
  • 12. Sharing of Study Data(Deidentified Individual-Patient Data, IPD)

    Sharing of Study Data Information - Sharing Statement, Time of Sharing, Way of Sharing
    Sharing Statement No
화면 최상단으로 이동

TOP

BOTTOM

화면 최하단으로 이동