ASSESSING A CHILD'S BONE AGE USING DEEP LEARNING ON HAND X-RAY IMAGES

Abstract

Intriduction. Assessing children's bone age is essential for detecting growth and developmental disorders. By comparing bone age to chronological age, physicians can identify deviations and diagnose issues.

Methods. In this study, we used a neural network model to predict the bone age of children based on X-ray images of their hands.

Results. The effectiveness of such application of the model is demonstrated in comparison with the average error results on the validation dataset.

Conclusions. Our study has demonstrated that the DenseNet169 architecture, when integrated with patient gender information, can effectively predict bone age from X-ray images, yielding competitive results compared to established models like Milvue. We hope this approach will significantly improve prediction accuracy and provide deeper insights into bone development. By focusing on critical anatomical regions and integrating relevant clinical data, we aim to set new standards in automated bone age assessment, ultimately benefiting pediatric healthcare.

References

Alhassan, M., & Mumuni, F. (2022). Data augmentation: A comprehensive survey of modern approaches. Array, 16, 100258. https://doi.org/10.1016/j.array.2022.100258

Greulich, W.W., & Pyle, S.I. (1959). Radiographic Atlas of Skeletal Development of the Hand and Wrist (2nd ed.). Stanford University Press. https://www.cambridge.org/core/journals/amg-acta-geneticae-medicae-et-gemellologiae-twin-research/article/w-w-greulich-and-s-i-pyle-radiographic-atlas-of-skeletal-development-of-the-hand-and-wrist-2nd-edition-i-volumeatlante-di-256-pagine-stanford-university-press-stanford-california-1959/40E97D3A609810F793A4228BE960A7CA

Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017). Densely connected convolutional networks. Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4700–4708). https:/doi.org/10.1109/CVPR.2017.243, https://www.researchgate.net/publication/306885833_Densely_Connected_Convolutional_Networks

Jabbar, A.J., & Abdulmunem, A. (2023). Bone age assessment based on deep learning architecture. International Journal of Electrical and Computer Engineering (IJECE), 13(2), 2078–2085. https:/doi.org/10.11591/ijece.v13i2.pp2078-2085

Lindner, C., Thiagarajah, S., Wang, C., Wilkinson, J. M., & Bowes, M. (2020). Bone age estimation using a deep learning model with anatomical domain knowledge. In Digital Image Computing: Techniques and Applications (DICTA), 2020 International Conference (pp. 610–617). Springer, Cham. https:/doi.org/10.1007/978-3-030-59725-2_61

Milvue. (n. d.). AI-powered medical imaging solutions. https://www.milvue.com/

Shorten, C., & Khoshgoftaar, T.M. (2019). A survey on image data augmentation for deep learning. Journal of Big Data, 6(1), 1–48. https://doi.org/10.1186/s40537-019-0197-0

Tan, M., & Le, Q.V. (2020). EfficientNet: Rethinking model scaling for convolutional neural networks. In International Conference on Machine Learning (pp. 6105–6114). PMLR. https://doi.org/10.48550/arXiv.1905.11946, https://proceedings.mlr.press/v97/tan19a.html