文献速递儿科影像骨肌DR

基于深度学习的小儿腹部X线片全自动Risser分期评估模型

Deep learning-based fully automatic Risser stage assessment model using abdominal radiographs.
PubMed 文献 · Pediatr Radiol · 2024-07-24

这项多中心研究回顾性收集三家医疗机构2019年1月至2022年4月间1681张仰卧位腹部X线片(年龄9-18岁,50%女性),采用美国Risser分期系统手动分级;利用DeepLabv3+骨盆骨分割模型提取双侧髂嵴图像块后训练ConvNeXt-B进行Risser分级,模型在内部测试集准确率0.87、AUROC 0.99、平均绝对误差0.13,外部验证集准确率0.75、AUROC 0.95、平均绝对误差0.26。

要点

  • 模型全自动完成骨盆分割与Risser分级,内部测试集准确率0.87、AUROC 0.99。
  • 外部验证集准确率0.75、AUROC 0.95、平均绝对误差0.26。
  • 分割模型使用90张骨盆CT数字重建X线片训练,分类模型基于ConvNeXt-B。

核心结论

  • 研究开发了基于深度学习的全自动分割与分类模型,可利用腹部X线片进行Risser分期评估。

研究设计与样本

  • 多中心回顾性研究,共纳入1681张仰卧位腹部X线片(年龄9-18岁,50%女性),来自三家医疗机构,采集时间为2019年1月至2022年4月;其中来自医院1和医院2的1577张图像用于模型开发,来自医院3的104张图像用于外部验证;以美国Risser分期系统手动分级为参考标准。

诊断标准与关键数值

  • 模型在内部测试集准确率0.87、平均绝对误差0.13、AUROC 0.99;在外部测试集准确率0.75、平均绝对误差0.26、AUROC 0.95。

临床应用与阅片要点

  • 原文未据此给出推荐。

局限性

  • 原文未报告局限性。

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

BACKGROUND: Artificial intelligence has been increasingly used in medical imaging and has demonstrated expert level performance in image classification tasks.

OBJECTIVE: To develop a fully automatic approach for determining the Risser stage using deep learning on abdominal radiographs.

MATERIALS AND METHODS: In this multicenter study, 1,681 supine abdominal radiographs (age range, 9-18 years, 50% female) obtained between January 2019 and April 2022 were collected retrospectively from three medical institutions and graded manually using the United States Risser staging system. A total of 1,577 images from Hospitals 1 and 2 were used for development, and 104 images from Hospital 3 for external validation. From each radiograph, right and left iliac crest patch images were extracted using the pelvic bone segmentation model DeepLabv3 + with the EfficientNet-B0 encoder trained with 90 digitally reconstructed radiographs from pelvic computed tomography scans with a pelvic bone mask. Using these patch images, ConvNeXt-B was trained to grade according to the Risser classification. The model's performance was evaluated using accuracy, area under the receiver operating characteristic curve (AUROC), and mean absolute error.

RESULTS: The fully automatic Risser stage assessment model showed an accuracy of 0.87 and 0.75, mean absolute error of 0.13 and 0.26, and AUROC of 0.99 and 0.95 on internal and external test sets, respectively.

CONCLUSION: We developed a deep learning-based, fully automatic segmentation and classification model for Risser stage assessment using abdominal radiographs.

来源:Pediatr Radiol,查看原文。版权归原作者所有。

引用

  1. Hwang JY, Kim Y, Hwang J, et al. Deep learning-based fully automatic Risser stage assessment model using abdominal radiographs. Pediatr Radiol. 2024. PMID: 39046527. doi:10.1007/s00247-024-05999-1

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。