Related Experiment Video
Updated: Jan 6, 2026

07:56
Scanning Skeletal Remains for Bone Mineral Density in Forensic Contexts
Published on: January 29, 2018
18.1K
Population-specific calibration and validation of an open-source bone age AI
Sebastian Rassmann1,2, Luka Abashishvili3, Elene Melikidze3
1Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Bonn, Germany.
Scientific Reports
|September 23, 2025
Summary
Artificial intelligence (AI) for bone age (BA) assessment was adapted for Georgian children. The calibrated AI model, Deeplasia-GE, significantly improved accuracy, offering a reliable tool for pediatric growth monitoring in this population.
Area of Science:
- Pediatric endocrinology
- Medical imaging analysis
- Artificial intelligence in healthcare
Background:
- Accurate bone age (BA) assessment is vital for monitoring pediatric growth and guiding treatments like hormonal therapy.
- Existing artificial intelligence (AI) models for BA assessment often lack generalizability due to training on limited, non-diverse datasets.
- Population-specific variations in bone growth necessitate tailored AI solutions for accurate skeletal maturity evaluation.
Purpose of the Study:
- To adapt and validate an open-source AI tool (Deeplasia) for bone age assessment in the Georgian pediatric population.
- To address concerns regarding the applicability of existing AI BA models to underrepresented ancestries.
- To improve the accuracy and reliability of automated BA assessment for Georgian children.
Main Methods:
- Retrospective collection of 381 pediatric hand X-rays from the Georgian population.
- Establishment of a manual bone age reference rating by seven local pediatric specialists.
- Sex-specific linear calibration of the Deeplasia AI model using 121 images, creating Deeplasia-GE.
- Evaluation of model performance on a held-out test set of 260 images.
Main Results:
- The default Deeplasia model showed a mean absolute difference (MAD) of 6.57 months, which improved to 5.69 months after calibration (Deeplasia-GE).
- Initial overestimation of BA by the default model (SMD of +2.85 months for females, +5.35 months for males) was significantly reduced by Deeplasia-GE (-0.03 months for females, +0.58 months for males).
- Deeplasia-GE demonstrated a smaller error than individual human raters and maintained high test-retest reliability.
Conclusions:
- The calibrated AI model, Deeplasia-GE, provides a reliable and accurate method for bone age assessment in Georgian children.
- This study highlights the importance of population-specific calibration for AI-driven medical diagnostic tools.
- Deeplasia-GE offers a valuable advancement for pediatric growth monitoring and treatment guidance within the Georgian healthcare system.
Related Concept Videos
Classification of Bones
9.4K
The bones of the human skeletal system are of varied shapes, sizes, and functions. They can be classified based on their shape and function into four major classes: long bones, short bones, flat bones, and irregular bones. Some classifications include a fifth type, the sesamoid bones, as a separate class, whereas others categorize them under short bones.
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The...
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The...
9.4K
Changes in the Appendicular Skeleton with Age
3.3K
The upper and lower limb initially develops as a small bulge called a limb bud, which appears on the lateral side of the early embryo. The upper limb bud appears near the end of the fourth week of development, with the lower limb bud appearing shortly after.
Initially, the limb buds consist of a core of mesenchyme covered by a layer of ectoderm. The ectoderm at the end of the limb bud thickens to form a narrow crest called the apical ectodermal ridge. This ridge stimulates the underlying...
Initially, the limb buds consist of a core of mesenchyme covered by a layer of ectoderm. The ectoderm at the end of the limb bud thickens to form a narrow crest called the apical ectodermal ridge. This ridge stimulates the underlying...
3.3K
Bone Remodeling
40.2K
Bone remodeling is a continuous and balanced process of bone resorption by osteoclasts and bone formation by osteoblasts. In adults, it helps maintain bone mass and calcium homeostasis. While mechanical stress can stimulate turnover as part of the normal maintenance and reparative process, several hormones also regulate bone remodeling.
40.2K

