Related Experiment Video
Updated: Oct 6, 2025
07:56
Scanning Skeletal Remains for Bone Mineral Density in Forensic Contexts
Published on: January 29, 2018
17.8K
Forensic bone age estimation of adolescent pelvis X-rays based on two-stage convolutional neural network
Li-Qin Peng1,2,3, Yu-Cheng Guo4,5, Lei Wan1
1Shanghai Key Laboratory of Forensic Medicine, Shanghai Forensic Service Platform, Academy of Forensic Science, 1347 GuangFu West Road, Shanghai, 200063, People's Republic of China.
International Journal of Legal Medicine
|January 18, 2022
Summary
Pelvic bone age estimation for teenagers is improved using AI segmentation networks to overcome X-ray image overlap issues. This method enhances accuracy in determining bone age for forensic and clinical applications.
Area of Science:
- Forensic Anthropology
- Medical Imaging Analysis
- Artificial Intelligence in Medicine
Background:
- Accurate bone age estimation is crucial in forensic science, particularly for adolescents.
- Pelvic X-rays are valuable for bone age assessment but often suffer from reduced accuracy due to overlapping pelvic organs.
- Existing methods struggle to precisely delineate key pelvic structures, limiting assessment reliability.
Purpose of the Study:
- To enhance the accuracy of pelvic bone age estimation in teenagers using deep learning segmentation.
- To automate the identification of critical pelvic regions, mitigating limitations of traditional X-ray analysis.
- To evaluate the performance improvement of convolutional neural networks (CNNs) after image segmentation.
Main Methods:
- Retrospective analysis of 2164 pelvic X-ray images from Chinese Han teenagers (11-21 years old).
- Utilized a U-Net segmentation network to detect and isolate key pelvic areas.
- Applied regional augmentation by combining segmented areas with original X-rays.
- Compared bone age estimation accuracy using three CNNs (Inception-V3, Inception-ResNet-V2, VGG19) with and without segmentation pre-processing.
Main Results:
- Segmentation significantly reduced Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) across all tested CNNs.
- RMSE decreased from 1.22-1.63 years to 0.93-1.14 years, and MAE from 0.93-1.23 years to 0.67-0.88 years post-segmentation.
- Visual comparisons using Bland-Altman plots and attention maps confirmed improved estimation consistency and accuracy.
Conclusions:
- AI-driven pelvic segmentation effectively overcomes X-ray image limitations, enhancing bone age estimation accuracy.
- The proposed method offers a more reliable approach for forensic and clinical bone age assessment in adolescents.
- U-Net segmentation combined with CNNs provides a robust framework for improving diagnostic precision in skeletal age determination.
Related Concept Videos
01:18Classification of Bones
8.0K
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...
8.0K
01:09Changes in the Appendicular Skeleton with Age
2.7K
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...
2.7K