Artificial Intelligence for breaST canceR scrEening in mAMmography (AI-STREAM)

Artificial Intelligence for breaST canceR scrEening in mAMmography (AI-STREAM): A Prospective, Multicenter Cohort Study

This prospective study aims to generate real-world evidence on the overall benefits and disadvantages of using Lunit INSIGHT MMG AI based CADe/x for breast cancer detection in a population-based breast cancer screening program in Korea.

Study Overview

• Status: Active, not recruiting
• Conditions: Breast Cancer
• Intervention / Treatment: Device: Lunit INSIGHT MMG CADe/x for medical imaging

Detailed Description

1. Several challenges have been identified in breast cancer screening: 1) Some breast cancer cases not identified through screening; 2) Excessive recalls for further testing; 3) Low sensitivity in dense breasts; 4) Inter-reader variability. AI-based CADe/x has been shown to improve radiologist performance and provides results equivalent or superior to those from radiologists alone.
2. This multicenter, prospective study involves women who visit sites for breast cancer screening in Korea. Women eligible for national cancer screening in the relevant year who read the study participant recruitment brochure and read and sign the Participant Information Sheet and Informed Consent Form will be recruited into this study. Approximately 32,714 participants will be enrolled from February 2021 through December 2022 at five study sites in Korea.
3. In Korea, a single radiologist performs mammogram readings. If recall is required (per usual care), further diagnostic work-up will be conducted to confirm cancer detected at screening. The national cancer registry databases will be reviewed in 2026 and 2027. Available findings will be recorded for all participants regardless of their screening status to identify study participants with breast cancer diagnosis within one year and within two years from screening.
4. In primary outcome measurement, as part of the standard screening procedure, mammograms will be read and recorded by a breast radiologist without AI-CADe/x, and then with AI-based CADe/x. [Set1]
5. In secondary outcome measurement, mammograms from the same participants as Set 1 will be read and recorded by a general radiologist without AI-based CADe/x, and then with AI-based CADe/x. [Set 2] In additional secondary outcome measurement, arbitration reading will be conducted by another breast radiologist without AI-based CADe/x for cases in which the reading results of the two radiologists without AI-based CADe/x in Set 1 and Set 2 are inconsistent. [Set 3]
6. After completing the standard screening procedure in Set 1, several situational comparison groups [Set2 and Set3] for comparison the diagnostic accuracy will be performed independently and retrospectively The results from Set 2 and Set 3 will not impact the clinical decision(s) associated with the care of the study participants.

• Study Type: Observational
• Enrollment (Actual): 25008

Contacts and Locations

Study Locations

• Korea, Republic of
  • Seongnam-si, Korea, Republic of
    • Department of Radiology, CHA bundang Medical Center
  • Seoul, Korea, Republic of, 04401
    • Department of Radiology, Soonchunhyang University Hospital
  • Seoul, Korea, Republic of
    • Department of Radiology, Konkuk University Medical Center
  • Seoul, Korea, Republic of
    • Department of Radiology, Kyung Hee University Hospital at Gangdong
  • Seoul, Korea, Republic of
    • Department of Radiology, Nowon Eulgi Medical center

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 40 years to 80 years (Adult, Older Adult)
• Accepts Healthy Volunteers: No

• Sampling Method: Probability Sample

Study Population

Participants of breast cancer screening in Korean women aged 40 years or older with average risk of breast cancer.

Description

Inclusion Criteria:

• Participants must meet all of the following inclusion criteria in order to be enrolled in the study:
• Be eligible for national cancer screening in the relevant year and visit the site for breast cancer screening
• Provide consent for study participation using the Informed Consent Form and complete a Participant information Sheet

Exclusion Criteria:

• Participants who meet any of the following criteria will be excluded from the study:
• Has a history of or current breast cancer
• Is currently pregnant or plans to become pregnant in the next 12 months
• Has a history of breast surgery (mammoplasty or insertion of a foreign substance, such as paraffin or silicon)
• Has mammography for diagnostic purposes

Study Plan

How is the study designed?

Design Details

• Observational Models: Cohort
• Time Perspectives: Prospective

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
same as study population; Use of AI-based CADe/x by breast radiologists	Device: Lunit INSIGHT MMG CADe/x for medical imaging; • A software that detects areas suspected of breast cancer using mammographic images, marks areas suspected of malignant lesions, and displays the probability of malignant lesions to assist with the interpreting physician's diagnosis

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
• Diagnostic accuracy with difference between breast radiologists with and without AI-based CADe/x	Diagnostic accuracy is assessed using cancer registry data as the reference group to calculate cancer detection rate [CDR], recall rate, sensitivity, positive predictive value	12months after screening, 24months after screening

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
• Diagnostic accuracy and difference of the following comparison groups with or without AI	Between general radiologist with and without AI-based CADe/x; Between breast radiologist arbitration reading and breast radiologist with AI-based CADe/x; Between breast radiologist arbitration reading and general radiologist with AI-based CADe/x; Between general radiologist with AI-based CADe/x and breast radiologist without AI-based CADe/x; Between breast radiologist without AI-based CADe/x and stand-alone AI-based CADe/x; Between general radiologist without AI-based CADe/x and stand-alone AI-based CADe/x; Between breast radiologist with AI-based CADe/x and general radiologist with AI-based CADe/x; Between breast radiologist without AI-based CADe/x and general radiologist without AI-based CADe/x Diagnostic accuracy is assessed using cancer registry data as the reference group to calculate CDR, recall rate, sensitivity, PPV, specificity, interval cancer rate, and AUROC.	12months after screening, 24 months after screening

Collaborators and Investigators

• Sponsor: Kyung Hee University Hospital at Gangdong
• Collaborators: Korea Health Industry Development Institute
• Investigators: Study Director: Yun-Woo Chang, MD, PhD, Soonchunhyang University Hospital, Seoul

Publications and helpful links

General Publications

• Kim HE, Kim HH, Han BK, Kim KH, Han K, Nam H, Lee EH, Kim EK. Changes in cancer detection and false-positive recall in mammography using artificial intelligence: a retrospective, multireader study. Lancet Digit Health. 2020 Mar;2(3):e138-e148. doi: 10.1016/S2589-7500(20)30003-0. Epub 2020 Feb 6.
• Salim M, Wahlin E, Dembrower K, Azavedo E, Foukakis T, Liu Y, Smith K, Eklund M, Strand F. External Evaluation of 3 Commercial Artificial Intelligence Algorithms for Independent Assessment of Screening Mammograms. JAMA Oncol. 2020 Oct 1;6(10):1581-1588. doi: 10.1001/jamaoncol.2020.3321.
• Chang YW, An JK, Choi N, Ko KH, Kim KH, Han K, Ryu JK. Artificial Intelligence for Breast Cancer Screening in Mammography (AI-STREAM): A Prospective Multicenter Study Design in Korea Using AI-Based CADe/x. J Breast Cancer. 2022 Feb;25(1):57-68. doi: 10.4048/jbc.2022.25.e4. Epub 2022 Jan 6. Erratum In: J Breast Cancer. 2022 Apr;25(2):147.

Study record dates

Study Major Dates

• Study Start (Actual): February 1, 2021
• Primary Completion (Actual): December 31, 2022
• Study Completion (Estimated): December 31, 2024

Study Registration Dates

• First Submitted: August 13, 2021
• First Submitted That Met QC Criteria: August 21, 2021
• First Posted (Actual): August 27, 2021

Study Record Updates

• Last Update Posted (Actual): September 28, 2023
• Last Update Submitted That Met QC Criteria: September 25, 2023
• Last Verified: September 1, 2023

More Information

Terms related to this study

• Keywords: breast cancer; screening; AI-based CADe/x
• Additional Relevant MeSH Terms: Skin Diseases; Neoplasms; Neoplasms by Site; Breast Diseases; Breast Neoplasms
• Other Study ID Numbers: oddie2

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No