An enhanced deep learning approach for brain cancer MRI images classification using residual networks

作者：

Highlights：

• We propose an enhanced deep learning approach for classifying brain tumor types from MRI images.

• Approach is using a benchmark dataset and has managed to achieve 99% accuracy which is state of the art result and higher than all previous work.

• Beside accuracy, we are using other metrics for evaluation such as precision, recall, f1-score and balanced accuracy to obtain accurate results against the imbalanced dataset.

摘要

•We propose an enhanced deep learning approach for classifying brain tumor types from MRI images.•Approach is using a benchmark dataset and has managed to achieve 99% accuracy which is state of the art result and higher than all previous work.•Beside accuracy, we are using other metrics for evaluation such as precision, recall, f1-score and balanced accuracy to obtain accurate results against the imbalanced dataset.

论文关键词：Machine learning,Artificial neural network,Convolutional neural network,Deep residual network,Cancer classification

论文评审过程：Received 22 July 2019, Revised 1 December 2019, Accepted 6 December 2019, Available online 10 December 2019, Version of Record 27 December 2019.

论文官网地址：https://doi.org/10.1016/j.artmed.2019.101779