NCT03530098

Validation of an Artificial Intelligence-based Algorithm for Skeletal Age Assessment

Completed NA Results posted Last updated 9 June 2021

What this trial tests

NA trial testing BoneAgeModel in Bone Age in 1,903 participants. Completed in 31 August 2019.

Timeline

12 July 2018

Primary endpoint
31 August 2019

31 August 2019

Quick facts

Lead sponsor	Stanford University
Phase	NA
Status	Completed
Study type	INTERVENTIONAL
Allocation	randomized
Design	parallel
Masking	none
Primary purpose	diagnostic
Enrollment	1,903
Start date	12 July 2018
Primary completion	31 August 2019
Estimated completion	31 August 2019
Sites	6 locations across United States

Drugs / interventions tested

• BoneAgeModel

Conditions studied

• Bone Age — all drugs for Bone Age →

Sponsor

Stanford University

Who can join

Eligibility, any sex, with Bone Age. Patients with the condition only — healthy volunteers not accepted.

Results — posted to ClinicalTrials.gov

Per-arm endpoint measurements with 95% confidence intervals where reported. Source: trial results section.

Paired Difference of Skeletal Age Estimate Primary · Up to 10 minutes to acquire the scan; up to 2 days to complete diagnosis review

Mean absolute difference between dictated final impressions (baseline measure by Radiologist) and the consensus determination of a panel of radiologists following review.

Group	Value	95% CI
Control (Without-AI)	5.95	5.53 – 6.37
Experiment (With-AI)	5.36	5.01 – 5.71

Time for Diagnosis Secondary · Up to approximately 4 minutes

Amount of time taken by radiologists when using the BoneAgeModel as compared to when they are not.

Group	Value	95% CI
Control (Without-AI)	142	80 – 248
Experiment (Without-AI)	102	59 – 196

Sponsor's own description

The purpose of this study is to understand the effects of using an Artificial Intelligence algorithm for skeletal age estimation as a computer-aided diagnosis (CADx) system. In this prospective real-time study, the investigators will send de-identified hand radiographs to the Artificial Intelligence algorithm and surface the output of this algorithm to the radiologist, who will incorporate this information with their normal workflows to make an estimation of the bone age. All radiologists involved in the study will be trained to recognize the surfaced prediction to be the output of the Artificial Intelligence algorithm. The radiologists' diagnosis will be final and considered independent to the output of the algorithm.

Publications & conference data

1 peer-reviewed publication reference this trial (live from Europe PMC):

1. Artificial Intelligence Algorithm Improves Radiologist Performance in Skeletal Age Assessment: A Prospective Multicenter Randomized Controlled Trial.
   Eng DK, Khandwala NB, Long J, Fefferman NR, et al · · 2021 · cited 51× · PMID 34581608 · DOI 10.1148/radiol.2021204021