Artificial Intelligence Assisted Bone Age Assessment

Comparison of the Difference in Bone Age Readings Among Radiologists Using Artificial Intelligence

Status

Enrolling by invitation

Phases

Unknown

Study type

Interventional

Source

ClinicalTrials.gov

Registry ID

NCT07121283

Enrollment

Registered

2025-08-13

Start date

2025-09-01

Completion date

2026-08-18

Last updated

2026-04-17

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Bone Age

Brief summary

AI-Assisted Bone Age Assessment

Detailed description

Artificial intelligence (AI) has gained great advancement in the application in clinical practice. However, this might introduce the automation bias, that the clinician over-rely on the incorrect advise from AI. The automation bias could have a great impact on the clinical decisions in an AI era. However, efforts to mitigate the automation bias tend to focus on upgrading AI performance and reducing bias in algorithms, which neglect the role of users. The aim of this study is to investigates the impact of the automation bias on the bone age assessment among radiologists with different seniority.

Interventions

DEVICETrue AI

True output from bone age AI

DEVICEFake AI

Randomized output from bone age AI

Study design

Allocation

RANDOMIZED

Intervention model

CROSSOVER

Primary purpose

DIAGNOSTIC

Masking

SINGLE (Subject)

Intervention model description

Six radiologists with different seniority (senior, intermediate, and junior) were recruited. The 200 standard films were evenly divided into two subsets: datasets A and B. A randomized crossover design with four periods (each with a 4-week duration) was arranged to minimize anticipation and carryover effects. Each radiologist who made an assessment with the aid of true AI or fake AI information could agree or disagree with the AI's predictions. The information of the fake AI was resulted from the randomization of the outputs from the true AI. Each radiologist was blinded to the existence of the fake AI.

Eligibility

Sex/Gender

ALL

Healthy volunteers

Inclusion criteria

* Exams read by radiologists who interpret pediatric skeletal age exams and verbally consent to participate

Exclusion criteria

* Exams containing more than one radiograph will not be included. No further

Design outcomes

Primary

Measure	Time frame	Description
Difference of Skeletal Age Estimate	Five months	Mean absolute difference of bone age assessment between true AI-assisted and fake AI-assisted among radiologists

Countries

Taiwan

Contacts

PRINCIPAL_INVESTIGATORYen-Huai Lin, Dr.

Cheng-Hsin General Hospital

Outcome results

None listed

Source: ClinicalTrials.gov · Data processed: Apr 18, 2026