AE-BoNet:使用未经监督的预训练模型进行儿科骨龄估计的深度学习方法
Mojtaba Sirati-Amsheh1, Elham Shabaninia2, Ali Chaparian1
1Department of Medical Physics, Faculty of Medicine, Isfahan University of Medical Sciences, Isfahan, Iran.
Journal of biomedical physics & engineering
|June 13, 2025
概括
这项研究引入了一种无监督的深度学习方法,用于使用自动编码器估计骨年龄,实现与当前方法相比精确的结果. 这种方法解决了放射图像分析中有限的标记数据所带来的挑战.
科学领域:
- 放射学 放射学是一门学科.
- 医疗成像医学成像
- 人工智能的人工智能
背景情况:
- 准确的骨龄评估对于评估生长和识别发育障碍至关重要.
- 骨年龄评估的传统临床方法耗时,并且受到观察者变化的影响.
- 深度学习为骨年龄估计提供了自动化解决方案,克服了手工技术的局限性.
研究的目的:
- 开发一种无监督的预训练方法,通过手部X射线图自动估计骨年龄.
- 为了应对骨龄估计中有限的标记数据的挑战.
- 利用自动编码器来学习放射图像中的独特特征,以改进估计.
主要方法:
- 利用北美放射学会 (RSNA) 的X射线图像收集进行评估.
- 训练了一种自动编码模型,用于重建手部放射图像.
- 采用基于训练有素编码器的模型进行最终的骨年龄估计.
主要成果:
- 在RSNAX射线图像采集上获得了9.3个月的平均绝对误差 (MAE).
- 已证明的性能与最先进的骨年龄估计方法相美.
- 验证了无监督预训练方法的有效性.
结论:
- 介绍了一种新的方法,用于用自动编码器进行未经监督的预训练,在手部放射图上估计骨龄.
- 突出了自动编码器和无监督学习作为传统方法的有效替代品的潜力.
- 展示了深度学习在推进骨年龄评估的自动化医学图像分析方面的重要性.
相关概念视频
01:40Bone Remodeling
41.3K
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.
41.3K
01:18Classification of Bones
15.9K
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...
15.9K