Pact-Net: Parallel CNNs and Transformers for medical image segmentation

•We propose a network with a dual-branch encoder for medical image segmentation tasks.•The proposed fusion module includes Channel-Spatial Fusion and Multi-Scale Fusion.•The Channel-Spatial Fusion is proposed to fuse features with the same resolution.•The Multi-Scale Fusion helps to reduce the semantic gap between encoder and decoder.•Our proposed method achieves competitive results in skin image segmentation tasks.