GasHis-Transformer: A multi-scale visual transformer approach for gastric histopathological image detection

Highlights：

• Considering the advantages of VT and CNN models, our GasHis-Transformer model integrates the describing capability of global and local information of VT’s and CNN’s.

• In GasHis-Transformer, the idea of multi-scale image analysis is introduced to describe the details of gastric tissues and cells under a microscope.

• GasHis-Transformer not only obtains good classification performance on gastric histopathological images but also shows an excellent generalization ability on other histopathological image datasets.

• A Dropconnect based lightweight network is proposed to reduce the model size and training time of GasHis-Transformer for clinical applications with improved confidence.