文献速递骨软组织肉瘤与黑色素瘤

深度学习卷积神经网络与58名皮肤科医生的皮肤镜黑色素瘤诊断性能对比

Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists.
PubMed 文献 · Ann Oncol · 2018-01-01

该比较性横断面读片研究使用100张皮肤镜图像测试集,对比Google Inception v4 CNN与58名国际皮肤科医生(含30名专家)对病变的二分诊断分类。CNN的ROC曲线下面积(AUC)为0.86,高于皮肤科医生平均0.79(P<0.01);在医生敏感度86.6%和88.9%水平上,CNN特异度82.5%均更高(P<0.01)。多数皮肤科医生表现不及CNN,提示医生可能受益于CNN图像分类辅助。

要点

  • CNN的ROC AUC为0.86,高于58名皮肤科医生平均0.79(P<0.01)。
  • 在医生敏感度86.6%和88.9%水平上,CNN特异度82.5%均更高(P<0.01)。
  • 多数皮肤科医生诊断表现不及CNN,无论经验多少均可能受益于CNN辅助。

核心结论

  • CNN对皮肤镜黑色素瘤的诊断分类性能优于多数皮肤科医生,医生可能受益于CNN图像分类辅助。

研究设计与样本

  • 比较性横断面读片研究,使用100张图像测试集,CNN与58名国际皮肤科医生(含30名专家)对比。
  • CNN为Google Inception v4架构,用皮肤镜图像及对应诊断训练和验证。
  • 读片分两级:level-I仅皮肤镜;level-II皮肤镜加临床信息与图像。

主要发现

  • level-I医生平均敏感度86.6%(±9.3%),特异度71.3%(±11.2%)。
  • level-II医生敏感度88.9%(±9.6%,P=0.19),特异度75.7%(±11.7%,P<0.05)。
  • CNN ROC AUC为0.86,高于医生平均ROC面积0.79(P<0.01)。
  • CNN在医生敏感度86.6%和88.9%时特异度均为82.5%,分别高于level-I的71.3%和level-II的75.7%(均P<0.01)。
  • CNN结果接近2016年ISBI挑战赛前三名算法。

关键数值与推荐

  • 主要结局为CNN与58名皮肤科医生对病变二分诊断分类的敏感度、特异度和ROC AUC。

临床怎么用

  • 皮肤科医生无论经验多少,均可考虑借助CNN图像分类辅助黑色素瘤诊断。

局限性

  • 原文摘要未报告局限性。

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

BACKGROUND: Deep learning convolutional neural networks (CNN) may facilitate melanoma detection, but data comparing a CNN's diagnostic performance to larger groups of dermatologists are lacking. METHODS: Google's Inception v4 CNN architecture was trained and validated using dermoscopic images and corresponding diagnoses. In a comparative cross-sectional reader study a 100-image test-set was used (level-I: dermoscopy only; level-II: dermoscopy plus clinical information and images). Main outcome measures were sensitivity, specificity and area under the curve (AUC) of receiver operating characteristics (ROC) for diagnostic classification (dichotomous) of lesions by the CNN versus an international group of 58 dermatologists during level-I or -II of the reader study. Secondary end points included the dermatologists' diagnostic performance in their management decisions and differences in the diagnostic performance of dermatologists during level-I and -II of the reader study. Additionally, the CNN's performance was compared with the top-five algorithms of the 2016 International Symposium on Biomedical Imaging (ISBI) challenge. RESULTS: In level-I dermatologists achieved a mean (±standard deviation) sensitivity and specificity for lesion classification of 86.6% (±9.3%) and 71.3% (±11.2%), respectively. More clinical information (level-II) improved the sensitivity to 88.9% (±9.6%, P = 0.19) and specificity to 75.7% (±11.7%, P < 0.05). The CNN ROC curve revealed a higher specificity of 82.5% when compared with dermatologists in level-I (71.3%, P < 0.01) and level-II (75.7%, P < 0.01) at their sensitivities of 86.6% and 88.9%, respectively. The CNN ROC AUC was greater than the mean ROC area of dermatologists (0.86 versus 0.79, P < 0.01). The CNN scored results close to the top three algorithms of the ISBI 2016 challenge. CONCLUSIONS: For the first time we compared a CNN's diagnostic performance with a large international group of 58 dermatologists, including 30 experts. Most dermatologists were outperformed by the CNN. Irrespective of any physicians' experience, they may benefit from assistance by a CNN's image classification. CLINICAL TRIAL NUMBER: This study was registered at the German Clinical Trial Register (DRKS-Study-ID: DRKS00013570; https://www.drks.de/drks_web/).

来源:Ann Oncol,查看原文。版权归原作者所有。

引用

  1. Haenssle HA, Fink C, Schneiderbauer R, et al. Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. Ann Oncol. 2018. PMID: 29846502. doi:10.1093/annonc/mdy166

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。