Comparison of atlas-based and artificial intelligence-based methods for bone age assessment in a pediatric population: agreement and influencing factors

Background Artificial intelligence (AI)-based tools are increasingly used for bone age assessment, but their agreement with traditional atlas-based methods requires further evaluation. Objective To compare bone age estimates obtained using the Greulich and Pyle Atlas, the Pocket Atlas of Skeletal Age, and the Hand Bone Age AI web application developed at Stanford University. Materials and methods A retrospective analysis was performed in 376 pediatric patients. Agreement between methods was assessed using intraclass correlation coefficient (ICC), Bland–Altman analysis, Friedman tests, and subgroup analyses according to age, sex, and pubertal status. Results Median bone age estimates were identical across methods (144 months). Overall agreement was high (ICC=0.93, 95% CI 0.92–0.94). Significant differences were observed between the AI-based method and both atlas-based methods, whereas no significant difference was found between the two atlas-based approaches. Clinical agreement within ±12 months reached 96.5% for the atlas-based methods and 77.4–83.8% for comparisons involving AI. Sex significantly influenced agreement, while pubertal status showed less consistent effects. Conclusion AI-based and atlas-based methods showed high overall agreement, although AI demonstrated greater variability at the individual patient level. Patient sex may influence agreement between assessment methods.

Authors

Publication Details

Journal
Pediatric Radiology
Published
2026-11-21
DOI
https://doi.org/10.1007/s00247-026-06852-3
Primary Topic
Forensic Anthropology and Bioarchaeology Studies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Comparison of atlas-based and artificial intelligence-based methods for bone age assessment in a pediatric population: agreement and influencing factors

Michał Dobrakowski, Marcin Ciekalski, Agnieszka Zachurzok, Marta Spruś et al.
Pediatric Radiology
Forensic Anthropology and Bioarchaeology Studies
article

Comparison of atlas-based and artificial intelligence-based methods for bone age assessment in a pediatric population: agreement and influencing factors

Michał Dobrakowski, Marcin Ciekalski, Agnieszka Zachurzok, Marta Spruś, Zuzanna Wantulok
article en

Abstract

Background Artificial intelligence (AI)-based tools are increasingly used for bone age assessment, but their agreement with traditional atlas-based methods requires further evaluation. Objective To compare bone age estimates obtained using the Greulich and Pyle Atlas, the Pocket Atlas of Skeletal Age, and the Hand Bone Age AI web application developed at Stanford University. Materials and methods A retrospective analysis was performed in 376 pediatric patients. Agreement between methods was assessed using intraclass correlation coefficient (ICC), Bland–Altman analysis, Friedman tests, and subgroup analyses according to age, sex, and pubertal status. Results Median bone age estimates were identical across methods (144 months). Overall agreement was high (ICC=0.93, 95% CI 0.92–0.94). Significant differences were observed between the AI-based method and both atlas-based methods, whereas no significant difference was found between the two atlas-based approaches. Clinical agreement within ±12 months reached 96.5% for the atlas-based methods and 77.4–83.8% for comparisons involving AI. Sex significantly influenced agreement, while pubertal status showed less consistent effects. Conclusion AI-based and atlas-based methods showed high overall agreement, although AI demonstrated greater variability at the individual patient level. Patient sex may influence agreement between assessment methods.

Pediatric Radiology
Forensic Anthropology and Bioarchaeology Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon