文献速递骨肌儿科影像技术与安全DR

手膝双输入深度学习评估青春期晚期骨龄

Dual-input deep learning bone age assessment in late adolescence using hand and knee radiographs.
PubMed 文献 · Skeletal Radiol · 2026-07-11

该单中心回顾性研究纳入547例13–19岁青少年,共1094张配对左手腕与左膝X线片。研究构建以ShuffleNet v2为基础、联合年龄与性别的双输入模型,按受试者划分训练、验证和内部测试集。内部测试显示平均绝对误差3.77个月、均方根误差4.58个月、组内相关系数0.984,98.18%预测值在参考标准12个月内。

要点

  • 98.18%的骨龄预测值与参考标准相差在12个月以内。
  • 膝X线应仅在已有或确有临床指征时使用,不作常规加拍。

核心结论

  • 该年龄辅助双输入框架在单中心回顾性队列中对青春期晚期骨龄评估显示出较好的内部性能。
  • 在临床部署、司法或争议年龄使用及法律决策前,需要独立多中心外部验证。
  • 膝X线片应仅在已经获得或确有医学必要时使用,不应作为常规额外影像检查。

研究设计与样本

  • 单中心回顾性研究,纳入547例13–19岁青少年,含配对左手腕和左膝X线片共1094张。
  • 数据按受试者层面分为训练集、验证集和内部留出测试集,比例为70%/10%/20%,对应382/55/110例受试者。
  • 模型以ShuffleNet v2为基础,将手和膝X线片作为视觉输入,并纳入实足年龄和性别作为辅助变量。
  • 验证集用于模型选择,最终性能在内部留出测试集上以平均绝对误差、均方根误差和组内相关系数评价。

关键影像表现/征象

  • 内部消融比较显示,手膝配对输入联合人口学变量的误差低于单一部位输入。

诊断标准与关键数值

  • 最终模型平均绝对误差为3.77个月,均方根误差为4.58个月,组内相关系数为0.984。
  • 98.18%的预测值与参考标准相差在12个月以内。

临床应用与阅片要点

  • 在日常阅片报告中,如遇青春期晚期骨龄评估,可关注手膝配对影像联合年龄性别的多源信息可能提升一致性。
  • 膝X线片不推荐仅为骨龄评估常规加拍,应结合已有影像和临床指征判断。

局限性

  • 该模型仅完成内部测试,缺乏独立多中心外部验证。
  • 研究为单中心回顾性设计,样本量和人群来源可能限制外推性。

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

OBJECTIVE: To develop and internally evaluate an age-assisted dual-input model for bone age assessment in late adolescence using paired hand-wrist and knee radiographs.

MATERIALS AND METHODS: This single-center retrospective study included 547 adolescents aged 13-19 years with paired left hand-wrist and left knee radiographs (1094 images) acquired during routine clinical care. Data were split at the subject level into training, validation, and internal held-out test sets (70%/10%/20%; 382/55/110 subjects), keeping paired images from each subject in the same subset. A ShuffleNet v2-based model used hand and knee radiographs as visual inputs and incorporated chronological age and sex as auxiliary variables. The validation set was used for model selection, and final performance was evaluated on the internal held-out test set using mean absolute error, root mean square error, and intraclass correlation coefficient.

RESULTS: The final model achieved a mean absolute error of 3.77 months, root mean square error of 4.58 months, and intraclass correlation coefficient of 0.984, with 98.18% of predictions within 12 months of the reference standard. Internal ablation comparisons showed lower errors with paired hand-knee inputs and demographic variables than with single-site inputs.

CONCLUSION: In this single-center retrospective cohort, the age-assisted dual-input framework showed promising internal performance for late-adolescent bone age assessment. Independent multicenter external validation is required before clinical deployment, forensic/disputed-age use, or legal decision-making. Knee radiographs should be used only when already available or medically justified, not as routine additional imaging.

来源:Skeletal Radiol,查看原文。版权归原作者所有。

引用

  1. Peng, Li, Wu, et al. Dual-input deep learning bone age assessment in late adolescence using hand and knee radiographs. Skeletal Radiol. 2026. PMID: 42435209. doi:10.1007/s00256-026-05297-x

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。