Automated classification of histopathology images using transfer learning

作者：

Highlights：

• Deep learning based ResNet-50 and DenseNet-161 pre-trained models employed to automatically classify histopathology images.

• The pre-trained CNN models are tested on color and grayscale histopathology images.

• The ResNet-50 and DenseNet-161 models outperform the existing studies to classify pathology patches into 24 categories.

• The proposed approaches can be used to reduce the workload of pathologists and improve diagnostic accuracy.

摘要

•Deep learning based ResNet-50 and DenseNet-161 pre-trained models employed to automatically classify histopathology images.•The pre-trained CNN models are tested on color and grayscale histopathology images.•The ResNet-50 and DenseNet-161 models outperform the existing studies to classify pathology patches into 24 categories.•The proposed approaches can be used to reduce the workload of pathologists and improve diagnostic accuracy.

论文关键词：Medical image classification,Histopathology,Deep learning,Transfer learning,CNN

论文评审过程：Received 9 August 2019, Revised 16 October 2019, Accepted 21 October 2019, Available online 3 November 2019, Version of Record 8 November 2019.

论文官网地址：https://doi.org/10.1016/j.artmed.2019.101743