性别因素重要吗?对市场批准的皮肤癌检测卷积神经网络诊断性能中性别相关差异的分析
该研究评估了市场批准的卷积神经网络(CNN)Moleanalyzer Pro®在皮肤癌分类中的性别相关偏差,筛查了开放获取皮肤镜图像库的性别分布,并在1549张图像(女性773张,男性776张)中测试诊断性能。结果显示,尽管训练数据中女性图像占比偏低(40% vs 男性60%),该CNN在女性和男性来源图像中的灵敏度、特异度和ROC-AUC均无统计学显著差异(p ≥ 0.59)。
要点
- 开放获取皮肤镜图像库中女性图像占比明显偏低(40% vs 男性60%)。
- 市场批准的CNN在女性与男性图像中的灵敏度分别为87.0%和87.1%,特异度分别为98.7%和96.9%。
- ROC-AUC在女性与男性图像中分别为0.984和0.979,性别差异无统计学意义(p ≥ 0.59)。
核心结论
- 尽管开放获取训练数据存在性别相关不平衡,所测试的CNN在皮肤病变分类中未显示性别相关偏差。
- 作者认为,医学人工智能算法的设计与训练应普遍考虑性别和社會性别维度。
研究设计与样本
- 研究筛查了广泛用于CNN训练的开放获取皮肤镜图像库中的性别分布。
- 在1549张按性别分层的皮肤镜图像中测试了市场批准的CNN(Moleanalyzer Pro®)的性别相关诊断性能,其中女性773张,男性776张。
主要发现
- 大多数开放获取图像库显示女性来源图像(40%)相较男性(60%)明显代表性不足。
- CNN在女性与男性来源图像中的灵敏度分别为87.0%[80.9%-91.3%]与87.1%[81.1%-91.4%]。
- CNN在女性与男性来源图像中的特异度分别为98.7%[97.4%-99.3%]与96.9%[95.2%-98.0%]。
- CNN在女性与男性来源图像中的ROC-AUC分别为0.984[0.975-0.993]与0.979[0.969-0.988]。
- 在本样本中,无论按图像分析还是额外的按个体分析,ROC-AUC的性别相关差异均无统计学意义(p ≥ 0.59)。
关键数值与推荐
- 女性来源图像灵敏度87.0%[80.9%-91.3%],男性来源图像灵敏度87.1%[81.1%-91.4%]。
- 女性来源图像特异度98.7%[97.4%-99.3%],男性来源图像特异度96.9%[95.2%-98.0%]。
- 女性来源图像ROC-AUC 0.984[0.975-0.993],男性来源图像ROC-AUC 0.979[0.969-0.988]。
- 按图像分析和按个体分析的性别相关ROC-AUC差异均无统计学意义(p ≥ 0.59)。
临床怎么用
- 原文未据此给出临床推荐。
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
BACKGROUND: Advances in biomedical artificial intelligence may introduce or perpetuate sex and gender discriminations. Convolutional neural networks (CNN) have proven a dermatologist-level performance in image classification tasks but have not been assessed for sex and gender biases that may affect training data and diagnostic performance. In this study, we investigated sex-related imbalances in training data and diagnostic performance of a market-approved CNN for skin cancer classification (Moleanalyzer Pro®, Fotofinder Systems GmbH, Bad Birnbach, Germany). METHODS: We screened open-access dermoscopic image repositories widely used for CNN training for distribution of sex. Moreover, the sex-related diagnostic performance of the market-approved CNN was tested in 1549 dermoscopic images stratified by sex (female n = 773; male n = 776). RESULTS: Most open-access repositories showed a marked under-representation of images originating from female (40%) versus male (60%) patients. Despite these imbalances and well-known sex-related differences in skin anatomy or skin-directed behaviour, the tested CNN achieved a comparable sensitivity of 87.0% [80.9%-91.3%] versus 87.1% [81.1%-91.4%], specificity of 98.7% [97.4%-99.3%] versus 96.9% [95.2%-98.0%] and ROC-AUC of 0.984 [0.975-0.993] versus 0.979 [0.969-0.988] in dermoscopic images of female versus male origin, respectively. In the sample at hand, sex-related differences in ROC-AUCs were not statistically significant in the per-image analysis nor in an additional per-individual analysis (p ≥ 0.59). CONCLUSION: Design and training of artificial intelligence algorithms for medical applications should generally acknowledge sex and gender dimensions. Despite sex-related imbalances in open-access training data, the diagnostic performance of the tested CNN showed no sex-related bias in the classification of skin lesions.
引用
- Sies K, Winkler JK, Fink C, et al. Does sex matter? Analysis of sex-related differences in the diagnostic performance of a market-approved convolutional neural network for skin cancer detection. Eur J Cancer. 2022. PMID: 35182926. doi:10.1016/j.ejca.2021.12.034
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。