文献速递泌尿生殖腹部MR

数据增强与迁移学习提高自动前列腺分割模型的泛化能力

Data Augmentation and Transfer Learning to Improve Generalizability of an Automated Prostate Segmentation Model.
PubMed 文献 · AJR Am J Roentgenol · 2020-10-14

本研究基于单中心2015年2月至2018年11月648例前列腺MRI回顾性队列,训练结合2D与3D架构及迁移学习的深度学习分割模型,并采用影像专用数据增强策略,在五个独立数据集(四个来自外部中心)上测试。迁移学习模型内部验证全前列腺平均Dice系数93.1、移行带89.0;外部测试中数据增强使全前列腺模型提升2.2%、移行带模型提升3.0%,而在测试中心数据上微调后效果最佳,全前列腺Dice系数91.5、移行带89.7。

要点

  • 迁移学习模型在内部验证集全前列腺分割平均Dice系数为93.1,移行带为89.0。
  • 仅用数据增强在外部测试集上使全前列腺模型Dice提升2.2%、移行带提升3.0%。
  • 在测试中心数据上微调后外部测试效果最佳,全前列腺Dice系数91.5、移行带89.7。

关键影像表现/征象

  • 应用迁移学习的前列腺分割模型在内部验证队列中全前列腺平均Dice系数为93.1、移行带为89.0。
  • 在多个测试集队列中,仅用数据增强使全前列腺模型性能提升2.2%、移行带分割模型提升3.0%。
  • 在测试中心数据上微调后测试集结果最佳,全前列腺分割平均Dice系数91.5、移行带89.7。

诊断标准与关键数值

  • 采用Dice相似系数评价模型分割性能。
  • 内部验证队列:全前列腺平均Dice系数93.1,移行带89.0。
  • 数据增强在测试集上的提升:全前列腺模型2.2%,移行带模型3.0%。
  • 微调后测试集最佳:全前列腺平均Dice系数91.5,移行带89.7。

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

OBJECTIVE. Deep learning applications in radiology often suffer from overfitting, limiting generalization to external centers. The objective of this study was to develop a high-quality prostate segmentation model capable of maintaining a high degree of performance across multiple independent datasets using transfer learning and data augmentation. MATERIALS AND METHODS. A retrospective cohort of 648 patients who underwent prostate MRI between February 2015 and November 2018 at a single center was used for training and validation. A deep learning approach combining 2D and 3D architecture was used for training, which incorporated transfer learning. A data augmentation strategy was used that was specific to the deformations, intensity, and alterations in image quality seen on radiology images. Five independent datasets, four of which were from outside centers, were used for testing, which was conducted with and without fine-tuning of the original model. The Dice similarity coefficient was used to evaluate model performance. RESULTS. When prostate segmentation models utilizing transfer learning were applied to the internal validation cohort, the mean Dice similarity coefficient was 93.1 for whole prostate and 89.0 for transition zone segmentations. When the models were applied to multiple test set cohorts, the improvement in performance achieved using data augmentation alone was 2.2% for the whole prostate models and 3.0% for the transition zone segmentation models. However, the best test-set results were obtained with models fine-tuned on test center data with mean Dice similarity coefficients of 91.5 for whole prostate segmentation and 89.7 for transition zone segmentation. CONCLUSION. Transfer learning allowed for the development of a high-performing prostate segmentation model, and data augmentation and fine-tuning approaches improved performance of a prostate segmentation model when applied to datasets from external centers.

来源:AJR Am J Roentgenol,查看原文。版权归原作者所有。

引用

  1. Sanford TH, Zhang L, Harmon SA, et al. Data Augmentation and Transfer Learning to Improve Generalizability of an Automated Prostate Segmentation Model. AJR Am J Roentgenol. 2020. PMID: 33052737. doi:10.2214/AJR.19.22347

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。