文献速递皮肤肿瘤

从互联网非标准化临床照片生成黑色素瘤与痣数据集

Generation of a Melanoma and Nevus Data Set From Unstandardized Clinical Photographs on the Internet.
PubMed 文献 · JAMA Dermatol · 2023-01-01

本诊断研究利用卷积神经网络等技术从约50万张互联网照片中半自动标注了5619张黑色素瘤或痣的裁剪病变图像(CAN5600)及2006张人工修订子集(CAN2000),并生成5000张合成图像(GAN5000)。在7个公开测试集上,基于CAN5600训练的EfficientNet Lite0的AUROC为0.874(0.042),高于基于现有公开数据集组合训练的0.809(0.063)。

要点

  • 从互联网非标准化照片半自动标注构建黑色素瘤与痣数据集CAN5600(n=5619)和CAN2000(n=2006)。
  • 基于CAN5600训练的CNN平均AUROC为0.874,高于现有公开数据集组合的0.809。
  • 合成数据集GAN5000(n=5000)训练的网络平均AUROC为0.838。

核心结论

  • 基于互联网图像创建的合成数据集可用于AI训练,且基于CAN5600训练的网络性能优于基于现有公开数据集组合训练的网络。
  • 标注数据集(CAN5600和LESION130k)与合成数据集(GAN5000)均可共享用于AI训练和医师间共识。

研究设计与样本

  • 诊断研究,从约50万张互联网照片中半自动标注5619张(CAN5600)和2006张(CAN2000)黑色素瘤或痣的裁剪病变图像。
  • 使用生成对抗网络(StyleGAN2-ADA)生成5000张合成图像(GAN5000),训练用CAN2000,预训练用LESION130k。
  • 测试使用7个现有公开数据集(共2312张图像),每次以其中一个作为测试集,比较EfficientNet Lite0 CNN在 proposed 数据集与其余6个现有数据集上的性能。

主要发现

  • 基于CAN5600训练的平均(SD)AUROC为0.874(0.042),P=0.02。
  • 基于现有数据集组合训练的平均(SD)AUROC为0.809(0.063)。

关键数值与推荐

  • 主要结局指标为判断恶性肿瘤的受试者工作特征曲线下面积(AUROC)。
  • EfficientNet Lite0 CNN在CAN5600、CAN2000、GAN5000上训练的AUROC均高于或等同于基于病理确诊的公开数据集训练的结果。

临床怎么用

  • 标注数据集(CAN5600和LESION130k)与合成数据集(GAN5000)可共享用于AI训练和医师间共识。
  • 基于互联网非标准化照片构建的数据集可用于训练 discriminative 和 generative AI。

局限性

  • 原文未述局限性

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

IMPORTANCE: Artificial intelligence (AI) training for diagnosing dermatologic images requires large amounts of clean data. Dermatologic images have different compositions, and many are inaccessible due to privacy concerns, which hinder the development of AI. OBJECTIVE: To build a training data set for discriminative and generative AI from unstandardized internet images of melanoma and nevus. DESIGN, SETTING, AND PARTICIPANTS: In this diagnostic study, a total of 5619 (CAN5600 data set) and 2006 (CAN2000 data set; a manually revised subset of CAN5600) cropped lesion images of either melanoma or nevus were semiautomatically annotated from approximately 500 000 photographs on the internet using convolutional neural networks (CNNs), region-based CNNs, and large mask inpainting. For unsupervised pretraining, 132 673 possible lesions (LESION130k data set) were also created with diversity by collecting images from 18 482 websites in approximately 80 countries. A total of 5000 synthetic images (GAN5000 data set) were generated using the generative adversarial network (StyleGAN2-ADA; training, CAN2000 data set; pretraining, LESION130k data set). MAIN OUTCOMES AND MEASURES: The area under the receiver operating characteristic curve (AUROC) for determining malignant neoplasms was analyzed. In each test, 1 of the 7 preexisting public data sets (total of 2312 images; including Edinburgh, an SNU subset, Asan test, Waterloo, 7-point criteria evaluation, PAD-UFES-20, and MED-NODE) was used as the test data set. Subsequently, a comparative study was conducted between the performance of the EfficientNet Lite0 CNN on the proposed data set and that trained on the remaining 6 preexisting data sets. RESULTS: The EfficientNet Lite0 CNN trained on the annotated or synthetic images achieved higher or equivalent mean (SD) AUROCs to the EfficientNet Lite0 trained using the pathologically confirmed public data sets, including CAN5600 (0.874 [0.042]; P = .02), CAN2000 (0.848 [0.027]; P = .08), and GAN5000 (0.838 [0.040]; P = .31 [Wilcoxon signed rank test]) and the preexisting data sets combined (0.809 [0.063]) by the benefits of increased size of the training data set. CONCLUSIONS AND RELEVANCE: The synthetic data set in this diagnostic study was created using various AI technologies from internet images. A neural network trained on the created data set (CAN5600) performed better than the same network trained on preexisting data sets combined. Both the annotated (CAN5600 and LESION130k) and synthetic (GAN5000) data sets could be shared for AI training and consensus between physicians.

来源:JAMA Dermatol,查看原文。版权归原作者所有。

引用

  1. Cho SI, Navarrete-Dechent C, Daneshjou R, et al. Generation of a Melanoma and Nevus Data Set From Unstandardized Clinical Photographs on the Internet. JAMA Dermatol. 2023. PMID: 37792351. doi:10.1001/jamadermatol.2023.3521

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。