融合迁移学习和数据增强的SC-Net模型在皮肤癌识别中的应用

打开文本图片集

关键词：皮肤癌诊断；DenseNet-201模型；XGBoost模型；特征融合；数据增强；注意力机制；少数类识别

中图分类号：TP391文献标志码：A

文章编号：1001-3695（2022）08-054-2550-06

doi：10.19734/j.issn.1001-3695.2021.12.0666

Application of SC-Net model integrated with transfer learning and data augmentation in skin cancer classification

Zuo Hangxu1，Liao Bin1，Chen Xiaokun1，Tong Yang2，Li Yong3

（1.College of Statistics & Data Science，Xinjiang University of Finance & Economics，Urumqi 830012，China；2.College of Electronic information，University of Electronic Science & Technology of China，Chengdu 610000，China；3.College of Integration of Traditional Chinese & Western Medicine，Southwest Medical University，Luzhou Sichuan 646100，China）

Abstract：The performance of the current skin cancer diagnostic models cannot meet the requirements of clinical applications，and the diagnostic accuracy is not high for a few categories.To solve this problem，this paper proposed a SC-Net model based on transfer learning and data augmentation.Firstly，it used the ECA attention module to fine tune the pre-training model of DenseNet-201 on the skin cancer dataset，and extracted the implicit high-level features of the images.Then，it joined the general statistical features，and used SMOTE oversampling technology to balance a few categories of data.Finally，it putt the data into XGBoost model for training to obtain the final SC-Net classification model.The experimental results show that the accuracy，sensitivity and specificity of SC-Net model reach 99.25%，99.25% and 99.88%，which is about 0.6%～18.7% higher than the existing models.The proposed model has stronger classification ability for a few categories such as Dermato fibroma and Actinic keratoses and intraepithelial carcinoma.

Key words：diagnosis of skin cancer；DenseNet-201 model；XGBoost model；feature fusion；data augmentation；attention mechanism；minority class identification

0引言

人类每年产生16.3 ZB的数据，到2025年，这一数字可能增加10倍[1]，特别伴随人口以及医疗需求的增加，医疗和医疗相关数据成为大数据时代不可或缺的组成部分。（剩余17758字）