Skin lesion segmentation using fully convolutional networks: A comparative experimental study
作者：
Highlights：
• A comparative study of lesion segmentation is shown by fully convolutional networks.
• FCN-AlexNet, FCN-8s, FCN-16s, and FCN-32s are applied on ISIC’17 for the first time.
• The Dice coefficients are proposed to compare successes of the used architectures.
• Detailed analysis on the success of the models is provided.
• The study has significant insights and can benefit researchers in future works.
摘要
•A comparative study of lesion segmentation is shown by fully convolutional networks.•FCN-AlexNet, FCN-8s, FCN-16s, and FCN-32s are applied on ISIC’17 for the first time.•The Dice coefficients are proposed to compare successes of the used architectures.•Detailed analysis on the success of the models is provided.•The study has significant insights and can benefit researchers in future works.
论文关键词：Deep Learning,Convolutional Neural Network,Fully Convolutional Network,Medical Image Segmentation
论文评审过程：Received 26 February 2020, Revised 15 May 2020, Accepted 9 July 2020, Available online 19 July 2020, Version of Record 24 July 2020.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.113742
