Momentum contrastive learning for few-shot COVID-19 diagnosis from chest CT images

作者：

Highlights：

• We formulate the COVID-19 diagnosis task as a few-shot learning problem.

• A self-supervised representation learning method is proposed to diagnose COVID-19 using only a limited number of samples.

• Our model is pre-trained on a general chest CT image dataset, andtested on two COVID-19 benchmarks. .

摘要

•We formulate the COVID-19 diagnosis task as a few-shot learning problem.•A self-supervised representation learning method is proposed to diagnose COVID-19 using only a limited number of samples.•Our model is pre-trained on a general chest CT image dataset, andtested on two COVID-19 benchmarks. .

论文关键词：COVID-19 diagnosis,Few-shot learning,Contrastive learning,Chest CT images

论文评审过程：Received 19 June 2020, Revised 13 November 2020, Accepted 22 November 2020, Available online 16 January 2021, Version of Record 19 January 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.107826