[论文解读] Monkeypox Skin Lesion Detection Using Deep Learning Models: A Feasibility Study

Shams Nafisa Ali, Md. Tazuddin Ahmed+5 more|arXiv (Cornell University)|Jul 6, 2022

Poxvirus research and outbreaks被引用 126

一句话总结

本文提出 Monkeypox Skin Lesion Dataset (MSLD)，并评估迁移学习的 CNNs（VGG16、ResNet50、InceptionV3）用于通过皮肤病变检测 monkeypox，在3折交叉验证中以 ResNet50 实现最佳准确率。还提出一个用于筛查的原型网页工具。

ABSTRACT

The recent monkeypox outbreak has become a public health concern due to its rapid spread in more than 40 countries outside Africa. Clinical diagnosis of monkeypox in an early stage is challenging due to its similarity with chickenpox and measles. In cases where the confirmatory Polymerase Chain Reaction (PCR) tests are not readily available, computer-assisted detection of monkeypox lesions could be beneficial for surveillance and rapid identification of suspected cases. Deep learning methods have been found effective in the automated detection of skin lesions, provided that sufficient training examples are available. However, as of now, such datasets are not available for the monkeypox disease. In the current study, we first develop the ``Monkeypox Skin Lesion Dataset (MSLD)" consisting skin lesion images of monkeypox, chickenpox, and measles. The images are mainly collected from websites, news portals, and publicly accessible case reports. Data augmentation is used to increase the sample size, and a 3-fold cross-validation experiment is set up. In the next step, several pre-trained deep learning models, namely, VGG-16, ResNet50, and InceptionV3 are employed to classify monkeypox and other diseases. An ensemble of the three models is also developed. ResNet50 achieves the best overall accuracy of $82.96(\pm4.57\%)$, while VGG16 and the ensemble system achieved accuracies of $81.48(\pm6.87\%)$ and $79.26(\pm1.05\%)$, respectively. A prototype web-application is also developed as an online monkeypox screening tool. While the initial results on this limited dataset are promising, a larger demographically diverse dataset is required to further enhance the generalizability of these models.

研究动机与目标

• 创建一个开放的 Monkeypox Skin Lesion Dataset (MSLD)，其中包含 monkeypox、chickenpox 和 measles 图像。
• 评估迁移学习 CNN 在二分类 monkeypox 与其他皮肤病变之间的可行性。
• 评估三种预训练模型（VGG16、ResNet50、InceptionV3）及其集成在准确性和一致性方面的表现。
• 基于这些模型开发一个用于远程 monkeypox 筛查的原型网络应用。

提出的方法

• 将 Monkeypox Skin Lesion Dataset (MSLD) 收集并预处理，有 228 张原始图像和 3192 张增强图像。
• 应用数据增强（旋转、平移、反射、色相、饱和度、亮度、噪声、模糊、缩放）。
• 通过解冻底部八层并添加带 dropout 的三层全连接（FC）层，对预训练 CNNs（VGG16、ResNet50、InceptionV3）进行微调。
• 使用 Adam 优化器（lr=1e-5）、二元交叉熵损失、批量大小 16，在 224x224x3 输入上进行 3 折交叉验证（70/10/20 拆分）进行训练。
• 使用准确率、精确率、F1 分数和敏感性进行评估；报告各折的均值和标准差。
• 将最佳模型部署到原型网络应用中，以对用户上传的病变进行分析。

实验结果

研究问题

• RQ1 公开的 Monkeypox Skin Lesion Dataset (MSLD) 能否实现基于机器学习的 monkeypox 从皮肤病变的筛查？
• RQ2 带有迁移学习的预训练 CNNs（VGG16、ResNet50、InceptionV3）能否在 MSLD 上实现可靠的二分类（monkeypox 与其他）？
• RQ3 这三种模型的集成在跨折之间是否更稳健或更一致？
• RQ4 使用这些模型的基于网络的筛查工具在远程初步评估方面的可行性如何？

主要发现

模型	准确率 (%)	精确率	召回率	F1 分数
VGG16	81.48 ± 6.87	0.85 ± 0.08	0.81 ± 0.05	0.83 ± 0.06
ResNet50	82.96 ± 4.57	0.87 ± 0.07	0.83 ± 0.02	0.84 ± 0.03
InceptionV3	74.07 ± 3.78	0.74 ± 0.02	0.81 ± 0.07	0.78 ± 0.04
Ensemble	79.26 ± 1.05	0.84 ± 0.05	0.79 ± 0.07	0.81 ± 0.02

• ResNet50 达到最高准确率：82.96% (±4.57)。
• VGG16 实现 81.48% (±6.87) 的准确率。
• InceptionV3 实现 74.07% (±3.78) 的准确率。
• Ensemble 实现 79.26% (±1.05) 的准确率，在跨折中拥有最低的标准差。
• 研究指出数据集规模有限和人口统计多样性不足是主要的泛化约束。
• 已开发一个用于在线筛查的原型网络应用。