文献速递皮肤肿瘤

使用深度学习算法对良性和恶性皮肤肿瘤临床图像进行分类

Classification of the Clinical Images for Benign and Malignant Cutaneous Tumors Using a Deep Learning Algorithm.
PubMed 文献 · J Invest Dermatol · 2018-02-08

本研究测试深度学习算法对12种皮肤病的临床图像进行分类,使用卷积神经网络(ResNet-152模型)以19,398张图像训练,并在Asan、Hallym和Edinburgh数据集上验证。在Asan数据集中,基底细胞癌、鳞状细胞癌、上皮内癌和黑色素瘤的曲线下面积分别为0.96±0.01、0.83±0.01、0.82±0.02和0.96±0.00;在Edinburgh数据集中对应疾病分别为0.90±0.01、0.91±0.01、0.83±0.01和0.88±0.01。在Hallym数据集中,基底细胞癌诊断的敏感性为87.1%±6.0%。该算法在480张Asan和Edinburgh图像上的表现与16名皮肤科医生相当。

要点

  • 卷积神经网络在Asan数据集中诊断基底细胞癌和黑色素瘤的AUC均为0.96。
  • 在Edinburgh数据集中,鳞状细胞癌的AUC为0.91,基底细胞癌为0.90。
  • 在Hallym数据集中,基底细胞癌诊断敏感性为87.1%±6.0%。

核心结论

  • 深度学习算法对多种皮肤肿瘤的临床图像分类表现出良好性能,在部分疾病中与皮肤科医生相当。

研究设计与样本

  • 使用卷积神经网络(Microsoft ResNet-152模型)进行图像分类,训练集来自Asan数据集训练部分、MED-NODE数据集和 atlas 网站图像,共19,398张图像;验证集来自Asan、Hallym和Edinburgh数据集的测试部分。

主要发现

  • 在Asan数据集中,基底细胞癌、鳞状细胞癌、上皮内癌和黑色素瘤的AUC分别为0.96±0.01、0.83±0.01、0.82±0.02和0.96±0.00。
  • 在Edinburgh数据集中,对应疾病的AUC分别为0.90±0.01、0.91±0.01、0.83±0.01和0.88±0.01。
  • 在Hallym数据集中,基底细胞癌诊断的敏感性为87.1%±6.0%。
  • 在480张Asan和Edinburgh图像上,算法性能与16名皮肤科医生相当。

关键数值与推荐

  • Asan数据集:基底细胞癌AUC 0.96±0.01,鳞状细胞癌AUC 0.83±0.01,上皮内癌AUC 0.82±0.02,黑色素瘤AUC 0.96±0.00。
  • Edinburgh数据集:基底细胞癌AUC 0.90±0.01,鳞状细胞癌AUC 0.91±0.01,上皮内癌AUC 0.83±0.01,黑色素瘤AUC 0.88±0.01。
  • Hallym数据集:基底细胞癌诊断敏感性87.1%±6.0%。

临床怎么用

  • 原文未据此给出推荐。

局限性

  • 为改善卷积神经网络性能,应收集更广年龄和种族范围的额外图像。

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

We tested the use of a deep learning algorithm to classify the clinical images of 12 skin diseases-basal cell carcinoma, squamous cell carcinoma, intraepithelial carcinoma, actinic keratosis, seborrheic keratosis, malignant melanoma, melanocytic nevus, lentigo, pyogenic granuloma, hemangioma, dermatofibroma, and wart. The convolutional neural network (Microsoft ResNet-152 model; Microsoft Research Asia, Beijing, China) was fine-tuned with images from the training portion of the Asan dataset, MED-NODE dataset, and atlas site images (19,398 images in total). The trained model was validated with the testing portion of the Asan, Hallym and Edinburgh datasets. With the Asan dataset, the area under the curve for the diagnosis of basal cell carcinoma, squamous cell carcinoma, intraepithelial carcinoma, and melanoma was 0.96 ± 0.01, 0.83 ± 0.01, 0.82 ± 0.02, and 0.96 ± 0.00, respectively. With the Edinburgh dataset, the area under the curve for the corresponding diseases was 0.90 ± 0.01, 0.91 ± 0.01, 0.83 ± 0.01, and 0.88 ± 0.01, respectively. With the Hallym dataset, the sensitivity for basal cell carcinoma diagnosis was 87.1% ± 6.0%. The tested algorithm performance with 480 Asan and Edinburgh images was comparable to that of 16 dermatologists. To improve the performance of convolutional neural network, additional images with a broader range of ages and ethnicities should be collected.

来源:J Invest Dermatol,查看原文。版权归原作者所有。

引用

  1. Han SS, Kim MS, Lim W, et al. Classification of the Clinical Images for Benign and Malignant Cutaneous Tumors Using a Deep Learning Algorithm. J Invest Dermatol. 2018. PMID: 29428356. doi:10.1016/j.jid.2018.01.028

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。