Huang Ying, Zhang Zhishuo, Ran Xin, Zhuang Kaiwen, Ran Yuping
Department of Dermatovenereology, West China Hospital, Sichuan University, Chengdu 610041, China.
School of Information and Software Engineering, University of Electronic Science and Technology of China (UESTC), Chengdu 610054, China.
Diagnostics (Basel). 2025 Aug 11;15(16):2011. doi: 10.3390/diagnostics15162011.
: Skin cancer diagnosis faces critical challenges due to the visual similarity of lesions and dataset limitations. : This study introduces HybridSkinFormer, a robust deep learning model designed to classify skin lesions from both clinical and dermatoscopic images. The model employs a two-stage architecture: a multi-layer ConvNet for local feature extraction and a residual-learnable multi-head attention module for global context fusion. A novel activation function (StarPRelu) and Enhanced Focal Loss (EFLoss) address neuron death and class imbalance, respectively. : Evaluated on a hybrid dataset (37,483 images across nine classes), HybridSkinFormer achieved state-of-the-art performance with an overall accuracy of 94.2%, a macro precision of 91.1%, and a macro recall of 91.0%, outperforming nine CNN and ViT baselines. : Its ability to handle multi-modality data and mitigate imbalance highlights its clinical utility for early cancer detection in resource-constrained settings.
由于病变的视觉相似性和数据集的局限性，皮肤癌诊断面临着严峻挑战。本研究引入了HybridSkinFormer，这是一种强大的深度学习模型，旨在从临床图像和皮肤镜图像中对皮肤病变进行分类。该模型采用两阶段架构：一个用于局部特征提取的多层卷积网络和一个用于全局上下文融合的残差可学习多头注意力模块。一种新颖的激活函数（StarPRelu）和增强型焦点损失（EFLoss）分别解决了神经元死亡和类别不平衡问题。在一个混合数据集（九个类别中的37483张图像）上进行评估时，HybridSkinFormer取得了领先的性能，总体准确率为94.2%，宏精度为91.1%，宏召回率为91.0%，优于九个卷积神经网络和视觉Transformer基线。其处理多模态数据和缓解不平衡的能力凸显了其在资源有限环境中早期癌症检测的临床实用性。
