文献速递口腔黏膜病

融合尺度不变特征变换的深度学习框架用于口腔黏膜病损分割

[Scale-invariant feature-enhanced deep learning framework for oral mucosal lesion segmentation].
PubMed 文献 · Zhonghua Kou Qiang Yi Xue Za Zhi · 2025-01-01

本研究基于838张口腔黏膜疾病临床白光图像构建PixelSIFT-UNet语义分割模型,按6∶2∶2分为训练集506张、验证集166张和测试集166张。以ResNet-50为骨干网络时Dice系数、mIoU、mPA和Precision分别为0.668、0.733、0.872和0.817,较传统U-Net和PSPNet模型均有显著提升。

要点

  • PixelSIFT-UNet可精确分割口腔扁平苔藓、口腔白斑病和口腔黏膜下纤维性变。
  • ResNet-50骨干时Dice 0.668,优于U-Net的0.662和PSPNet的0.671。
  • 各性能指标较传统语义分割模型均有显著提升。

核心结论

  • PixelSIFT-UNet在口腔黏膜病损分割任务中表现卓越,性能指标显著优于传统语义分割模型,可为提高分割精度提供有效工具。

研究设计与样本

  • 纳入浙江大学医学院附属口腔医院2020年1月至2022年12月收集的838张口腔黏膜疾病临床白光图像,按6∶2∶2随机分为训练集506张、验证集166张和测试集166张。
  • 采用Labelme软件标注病损边界,基于PixelSIFT-UNet构建深度学习模型,分别以VGG-16与ResNet-50为骨干网络,通过验证集优化参数并在测试集评估。

主要发现

  • PixelSIFT-UNet能精确分割口腔扁平苔藓、口腔白斑病和口腔黏膜下纤维性变3种常见口腔黏膜疾病。
  • VGG-16骨干时Dice系数、mIoU、mPA和Precision分别为0.642、0.699、0.836和0.792。
  • ResNet-50骨干时Dice系数、mIoU、mPA和Precision分别为0.668、0.733、0.872和0.817。
  • 传统U-Net模型相应指标为0.662、0.717、0.861和0.809;PSPNet模型相应指标为0.671、0.721、0.858和0.813。
  • PixelSIFT-UNet各性能指标较传统U-Net和PSPNet模型均有显著提升。

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)。

原文

Objective: To develop PixelSIFT-UNet, a novel semantic segmentation model that integrates deep learning with scale-invariant feature transform (SIFT) algorithm to improve the segmentation accuracy of oral mucosal lesions. Methods: This investigation utilized 838 standard clinical white light images of oral mucosal diseases acquired from January 2020 to December 2022 at the Stomatology Hospital Zhejiang University School of Medicine. Randomization was achieved through Python's random.seed function implementation. The random sample function was subsequently applied for sampling distribution. The dataset was stratified into three subsets with a 6∶2∶2 ratio: training (n=506), validation (n=166), and testing (n=166). Lesion boundaries were annotated using Labelme software, and a PixelSIFT-UNet-based deep learning model was developed with VGG-16 and ResNet-50 backbone networks. Model parameters were optimized using the validation set, and performance metrics [including Dice coefficient, mean intersection over union (mIoU), mean pixel accuracy (mPA), and Precision] were assessed on the test set. The model's performance was benchmarked against conventional semantic segmentation frameworks (U-Net and PSPNet). Results: The developed PixelSIFT-UNet model could achieve precise segmentation of three common oral mucosal lesions: oral lichen planus, oral leukoplakia, and oral submucous fibrosis. Utilizing VGG-16 as the backbone network, the model achieved Dice coefficient, mIoU, mPA, and Precision values of 0.642, 0.699, 0.836, and 0.792, respectively. Implementation with ResNet-50 backbone network yielded metrics of 0.668, 0.733, 0.872 and 0.817, demonstrating significant improvements across all performance indicators compared to conventional U-Net model (relevant metrics: 0.662, 0.717, 0.861 and 0.809) and PSPNet model (relevant metrics: 0.671, 0.721, 0.858 and 0.813). Conclusions: The proposed PixelSIFT-UNet architecture demonstrates superior performance in oral mucosal lesion segmentation tasks, surpassing conventional semantic segmentation models and providing robust quantitative improvements in segmentation accuracy. 目的: 开发一种融合深度学习与尺度不变特征变换算法的口腔黏膜病损语义分割模型(PixelSIFT-UNet),以提高口腔黏膜疾病病损区域分割精度。 方法: 纳入浙江大学医学院附属口腔医院2020年1月至2022年12月期间收集的838张符合标准的口腔黏膜疾病临床白光图像,使用Python语言random.seed函数设置随机种子,通过random.sample函数进行随机抽样,按6∶2∶2的比例划分为训练集(506张)、验证集(166张)和测试集(166张)。采用Labelme软件对训练集图像进行病损边界标注,并基于PixelSIFT-UNet构建深度学习模型,分别使用VGG-16与ResNet-50作为骨干网络。通过验证集优化模型参数,在测试集上评估模型性能,包括Dice系数、平均交并比(mIoU)、平均像素准确率(mPA)和精确率(Precision),并与传统语义分割模型(U-Net、PSPNet)进行对比分析。 结果: 所开发的PixelSIFT-UNet模型能精确分割口腔扁平苔藓、口腔白斑病和口腔黏膜下纤维性变3种常见的口腔黏膜疾病。采用VGG-16为骨干网络时,Dice系数、mIoU、mPA和Precision分别达到0.642、0.699、0.836和0.792;采用ResNet-50为骨干网络时,相应指标分别达到0.668、0.733、0.872和0.817,较传统U-Net模型(相应指标:0.662、0.717、0.861和0.809)、PSPNet模型(相应指标:0.671、0.721、0.858和0.813)的各项性能指标均有显著提升。 结论: 所开发的PixelSIFT-UNet模型在口腔黏膜病损分割任务中表现卓越,其性能指标显著优于传统语义分割模型,可为提高口腔黏膜病损分割精度提供有效工具。.

来源:Zhonghua Kou Qiang Yi Xue Za Zhi,查看原文。版权归原作者所有。

引用

  1. Zhang R, Jin L, Chen QM, et al. [Scale-invariant feature-enhanced deep learning framework for oral mucosal lesion segmentation]. Zhonghua Kou Qiang Yi Xue Za Zhi. 2025. PMID: 40015705. doi:10.3760/cma.j.cn112144-20241210-00467

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。