医生模拟器:Delta-Age-Sex-AdaIn 通过 AdaIn 风格迁移增强骨龄评估
研究提出 DASA-net 模型,将年龄与性别分布通过自适应实例归一化(AdaIN)融入风格迁移,模拟与标准骨龄图谱的视觉比对过程进行骨龄评估。在公开 RSNA 儿童骨龄数据集(14,236 张手部 X 线片,1–228 月龄)和自贡市第四人民医院私有数据集(474 张手部 X 线片,12–218 月龄,268 例男性)上,DASA-net 的平均绝对偏差(MAD)分别为 3.52 个月和 3.82 个月,优于 BoneXpert、Deeplasia、BoNet 等其他方法。
要点
- DASA-net 在 RSNA 数据集上 MAD 为 3.52 个月,优于 BoneXpert、Deeplasia、BoNet 等对比方法。
- 在自贡市第四人民医院私有数据集上 MAD 为 3.82 个月,同样优于其他方法。
- 消融实验用于证明整合年龄分布与性别分布的必要性。
核心结论
- DASA-net 通过将年龄和性别分布整合进风格迁移,帮助模型学习不同年龄和性别手部骨骼的特征,从而提升骨龄评估准确性。
研究设计与样本
- 研究提出 DASA-net 模型,由 BoneEncoder、Binary code distribution、Delta-Age-Sex-AdaIn 和 AgeDecoder 四个模块组成。在公开 RSNA 儿童骨龄预测数据集(14,236 张手部 X 线片,1–228 月龄)和自贡市第四人民医院私有骨龄预测数据集(474 张手部 X 线片,12–218 月龄,268 例男性)上与当前先进方法进行比较,并设计消融实验以证明整合年龄分布与性别分布的必要性。
诊断标准与关键数值
- DASA-net 在 RSNA 数据集上平均绝对偏差(MAD)为 3.52 个月,优于 BoneXpert、Deeplasia、BoNet 及其他深度学习方法。
- DASA-net 在自贡市第四人民医院私有数据集上 MAD 为 3.82 个月,同样优于其他方法。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
BACKGROUND: Bone age assessment assists physicians in evaluating the growth and development of children. However, deep learning methods for bone age estimation do not currently incorporate differential features obtained through comparisons with other bone atlases.
OBJECTIVE: To propose a more accurate method, Delta-Age-Sex-AdaIn (DASA-net), for bone age assessment, this paper combines age and sex distribution through adaptive instance normalization (AdaIN) and style transfer, simulating the process of visually comparing hand images with a standard bone atlas to determine bone age.
MATERIALS AND METHODS: The proposed Delta-Age-Sex-AdaIn (DASA-net) consists of four modules: BoneEncoder, Binary code distribution, Delta-Age-Sex-AdaIn, and AgeDecoder. It is compared with state-of-the-art methods on both a public Radiological Society of North America (RSNA) pediatric bone age prediction dataset (14,236 hand radiographs, ranging from 1 to 228 months) and a private bone age prediction dataset from Zigong Fourth People's Hospital (474 hand radiographs, ranging from 12 to 218 months, 268 male). Ablation experiments were designed to demonstrate the necessity of incorporating age distribution and sex distribution.
RESULTS: The DASA-net model achieved a lower mean absolute deviation (MAD) of 3.52 months on the RSNA dataset, outperforming other methods such as BoneXpert, Deeplasia, BoNet, and other deep learning based methods. On the private dataset, the DASA-net model obtained a MAD of 3.82 months, which is also superior to other methods.
CONCLUSION: The proposed DASA-net model aided the model's learning of the distinctive characteristics of hand bones of various ages and both sexes by integrating age and sex distribution into style transfer.
引用
- Wang L, Zhang X, Chen P, et al. Doctor simulator: Delta-Age-Sex-AdaIn enhancing bone age assessment through AdaIn style transfer. Pediatr Radiol. 2024. PMID: 39060414. doi:10.1007/s00247-024-06000-9
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。