An unsupervised feature learning framework for basal cell carcinoma image analysis

作者：

Highlights：

• A framework for basal cell carcinoma detection based on unsupervised feature learning.

• Experimental results show an improvement when compared to state-of-the-art methods.

• The framework integrates a digital staining method which improves interpretability.

• Digital staining highlights regions in the image which the model relates to cancer.

摘要

Highlights•A framework for basal cell carcinoma detection based on unsupervised feature learning.•Experimental results show an improvement when compared to state-of-the-art methods.•The framework integrates a digital staining method which improves interpretability.•Digital staining highlights regions in the image which the model relates to cancer.

论文关键词：Digital pathology,Representation learning,Unsupervised feature learning,Basal cell carcinoma

论文评审过程：Received 1 October 2014, Revised 9 April 2015, Accepted 15 April 2015, Available online 23 April 2015, Version of Record 9 June 2015.

论文官网地址：https://doi.org/10.1016/j.artmed.2015.04.004