Dataset Overview
Definition：Bone Age X-Ray Imaging Datasetis provided byChangsha Langhui Information Technology Co., Ltd.The database of images dedicated to the assessment of the bone age of children and young people is built by Long Salang Langhui Information Technology Ltd. The primary part is a positive x-ray of the left wrist, which preserves original DICOM images and exposure parameters, with corresponding bone values, a version of the assessment methodology and endocrine clinical background data, covering normal age references, early bone age and delayed full spectrum distribution of bone age.
| Dataset Name | Bone Age X-Ray Imaging Dataset |
| Total Data Volume | Total X-Ray Volume Included: 1,700,000 Cases |
| Imaging Modality | X-Ray (DR/CR) |
| Covered Body Parts | Standard positive left wrist, covering the finger, the palm, the wrist and the long end of the sept. |
| Raw Data | Dissensitization original DICOM or device raw digital image, with exposure parameters, pixel spacing, position/view maintained |
| Assessment methodology | Greulich-Pyle (GP) / Tanner-Whitehouse (TW3) Double Standard Compatibility |
| Normal reference | Not applicable 90% positive; normal age reference, early and late bone age must be covered |
| Source Institution | Cooperative TAC/Child Endocrinology Section |
| Quality Assurance | Radiologists mark + Deputy Director Medical Officers examine, double-mass control |
| Use Cases | Child bone age assessment AI, assisted diagnosis of endocrine diseases, growth and development monitoring |
Delivery Specifications and Field Requirements
The data set strictly follows the XR-BONE-AGE uniform delivery index requirements to ensure data quality and traceability.
Delivery & Counting：Line corresponds to one bone age check/Study; multiseries/multiview only 1 case; weight by primary key
Coverage：Mainly with the left wrist standard; other sideding or method to be specified
Positives and Composition：90% positive requirement not applied; normal age reference, early and delayed bone age are covered, by gender and age continuum
Raw Data Requirements：D-identified orifice DICOM, retracing exposure parmesis, pixel working, body position/view, deviance manufacter and model and original/treatment Status
Protocol and View:Right left wrist; original image separated from treatment; automatic bone age derivative separately marked
Cross-Data Association：Priority associated growth curves, endocrinological tests and follow-up; automatic bone age results should not be impersonated as a doctor ' s final report
2026 Frontier AI Research Progress
Domain Review：The OSA is evolving from the GP mapping method to the TIW3 deep learning model, with a focus light quantitative model, adaptation of Chinese population and bio-knowledge orientation in 2025-2026, with an accuracy rate of over 81.5%.
TW3 Advanced Learning Bonest Age Forecast Model
Diagnostics, 2025Based on the depth learning model of the Tanner-Whitehouse 3 method, the accuracy rate was 81.5%, based on the use of 560 Korean children's wrist DR tablets compared to paediatric radiologists.
BoneView BoneAge commercial AI software
Proof-of-concept, 2025The comparison of the GP mapping and TW2 methodology, which included 203 paediatric patients (average age 9.0), validated the clinical applicability of the commercial AI Bone Age software.
Light Quantification of Efficient NetB3 Chinese Children ' s Bones Age
2025Improvement of EfficientNetB3 architecture, R Adam Optimizer and Group Loss Functions, with only 15.8 million parameters, test set accuracy rate of 81.5%, suitable for the Chinese population of children.
Bioknowledge leads to a double-neurological bone-age projection.
Frontiers in Radiology, 2026The bi-neurological network framework, which combines anatomy and developmental knowledge, integrates biological aforethought into the skeletal age projection, enhancing model interpretability and generalization.
AIBA-GP Children and Adolescents ' Bonest Age Model in China
2025Based on the construction of RRSNA 14,236 left-hand DR tablets, the clinical applicability of children and adolescents in China is assessed, covering a continuous distribution of many ages.
Annotation Workflow & Quality Control
Image Acquisition & De-identification
Standardized acquisition workflow: complete data is exported directly from the equipment, and patient identifying information (PHI) is removed before the data is stored。
Initial Annotation (Specialist Physician)
Attending physicians annotate case by case against the standards, including lesion localization, morphological description and disease term determination。
Review (Associate Chief Physician or Above)
Experts with the title of Associate Chief Physician or above review the initial annotation results item by item, correcting erroneous annotations and supplementing missing dimensions。
Consistency Assessment
10% of samples are randomly selected and independently annotated by 3 physicians to calculate Fleiss' Kappa; below 0.Dimensions scored 75 are flagged for rework。
License and Usage Agreement
Academic Research License
For universities and research institutions, supporting academic research related to medical AI。After signing the agreement, a de-identified data subset is provided; the source must be acknowledged：Langhui Technology DataAssetsAPI。
Commercial License
For medical device companies and AI diagnostics companies, supporting medical AI product R&D and medical device registration。Provides full datasets + custom annotation + incremental update services。
Data Compliance Statement
All images come from legally authorized sources; PHI fields are fully de-identified and contain no information that can directly identify an individual, in compliance with the Personal Information Protection Law and the Data Security Law。
Customized services
Supports extended requirements such as adding specific disease types, multimodal annotation expansion, and custom training of AI-assisted diagnostic models。
Quick Facts
AI Frontier Research
The OSA is evolving from the GP mapping method to the TIW3 deep learning model, with a focus light quantitative model, adaptation of Chinese population and bio-knowledge orientation in 2025-2026, with an accuracy rate of over 81.5%.