面部与头皮皮损:市场批准的卷积神经网络与64位皮肤科医生的分类对比
本研究用市场批准的CNN(Moleanalyzer-Pro)对100张面部与头皮皮损的皮肤镜图像进行二分类,并与64位皮肤科医生(分为仅皮肤镜与结合临床信息两个级别)的管理决策比较。在固定特异性为医生平均水平(69.4%)时,CNN敏感度(96.2%[87.0%-98.9%])显著高于医生(84.2%[82.2%-86.2%],p<0.001),各培训级别医生均被超越;外部数据集中该表现大体得到验证,但在一组澳大利亚重度日光损伤皮肤数据中特异性降低。
要点
- 在同等特异性下,CNN敏感度显著高于64位皮肤科医生的管理决策。
- 各培训级别皮肤科医生的诊断表现均被该CNN超越(均p<0.001)。
- 外部验证大体支持该结果,但重度日光损伤皮肤数据中特异性下降。
核心结论
- 在同等特异性下,市场批准的CNN对面部与头皮皮损的敏感度显著高于64位皮肤科医生,作为辅助系统可能改善皮肤癌的早期检出。
研究设计与样本
- 采用市场批准的CNN(Moleanalyzer-Pro, FotoFinder Systems)对100张面部与头皮皮损的皮肤镜图像进行二分类。
- 同一批皮损用于两级阅读者研究,共纳入64位皮肤科医生(I级:仅皮肤镜;II级:皮肤镜、临床特写图像与文本信息)。
- 主要终点为CNN与医生II级管理决策的敏感度和特异性对比,并用四个额外外部数据集检验泛化性。
主要发现
- CNN的敏感度、特异度与ROC AUC分别为96.2%[87.0%-98.9%]、68.8%[54.7%-80.1%]与0.929[0.880-0.978]。
- II级中皮肤科医生管理决策的平均敏感度为84.2%[82.2%-86.2%],特异度为69.4%[66.0%-72.8%]。
- 将CNN特异度固定为医生平均特异度69.4%时,CNN敏感度(96.2%[87.0%-98.9%])显著高于医生(84.2%[82.2%-86.2%];p<0.001)。
- 各培训级别医生均被CNN超越(均p<0.001);CNN准确率(83.0%)显著高于医生II级管理决策的准确率(均p<0.001)。
- CNN表现在三个额外外部数据集中大体得到证实,但在一组包含重度日光损伤皮肤面部与头皮皮损的澳大利亚数据集中特异度尤其降低。
关键数值与推荐
- CNN敏感度96.2%[87.0%-98.9%]、特异度68.8%[54.7%-80.1%]、ROC AUC 0.929[0.880-0.978]。
- 医生II级管理决策平均敏感度84.2%[82.2%-86.2%]、特异度69.4%[66.0%-72.8%]。
- 固定CNN特异度为69.4%时,CNN敏感度96.2%[87.0%-98.9%]显著高于医生84.2%[82.2%-86.2%](p<0.001)。
- CNN准确率83.0%,显著高于医生II级管理决策准确率(均p<0.001)。
临床怎么用
- 该CNN可作为辅助系统用于面部与头皮皮损的皮肤镜分类,在同等特异度下提高敏感度以助早期检出皮肤癌。
- 在重度日光损伤皮肤人群中应用时需注意特异度可能下降,应结合临床信息判断。
局限性
- 外部验证中一组澳大利亚重度日光损伤皮肤数据集显示CNN特异度降低,泛化性存在人群差异。
- 原文未报告医生II级管理决策的具体准确率数值及外部数据集样本量等细节。
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: The clinical differentiation of face and scalp lesions (FSLs) is challenging even for trained dermatologists. Studies comparing the diagnostic performance of a convolutional neural network (CNN) with dermatologists in FSL are lacking. METHODS: A market-approved CNN (Moleanalyzer-Pro, FotoFinder Systems) was used for binary classifications of 100 dermoscopic images of FSL. The same lesions were used in a two-level reader study including 64 dermatologists (level I: dermoscopy only; level II: dermoscopy, clinical close-up images, textual information). Primary endpoints were the CNN's sensitivity and specificity in comparison with the dermatologists' management decisions in level II. Generalizability of the CNN results was tested by using four additional external data sets. RESULTS: The CNN's sensitivity, specificity and ROC AUC were 96.2% [87.0%-98.9%], 68.8% [54.7%-80.1%] and 0.929 [0.880-0.978], respectively. In level II, the dermatologists' management decisions showed a mean sensitivity of 84.2% [82.2%-86.2%] and specificity of 69.4% [66.0%-72.8%]. When fixing the CNN's specificity at the dermatologists' mean specificity (69.4%), the CNN's sensitivity (96.2% [87.0%-98.9%]) was significantly higher than that of dermatologists (84.2% [82.2%-86.2%]; p < 0.001). Dermatologists of all training levels were outperformed by the CNN (all p < 0.001). In confirmation, the CNN's accuracy (83.0%) was significantly higher than dermatologists' accuracies in level II management decisions (all p < 0.001). The CNN's performance was largely confirmed in three additional external data sets but particularly showed a reduced specificity in one Australian data set including FSL on severely sun-damaged skin. CONCLUSIONS: When applied as an assistant system, the CNN's higher sensitivity at an equivalent specificity may result in an improved early detection of face and scalp skin cancers.
引用
- Haenssle HA, Winkler JK, Fink C, et al. Skin lesions of face and scalp - Classification by a market-approved convolutional neural network in comparison with 64 dermatologists. Eur J Cancer. 2020. PMID: 33370644. doi:10.1016/j.ejca.2020.11.034
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。