Papers › HiFuse: Hierarchical Multi-Scale Feature Fusion Network for Medical Image Classification

HiFuse: Hierarchical Multi-Scale Feature Fusion Network for Medical Image Classification

21 Sep 2022arXiv:2209.10218archive 2025-07-28

Xiangzuo Huo, Gang Sun, Shengwei Tian, Yan Wang, Long Yu, Jun Long, Wendong Zhang, Aolun Li

Medical image classification has developed rapidly under the impetus of the convolutional neural network (CNN). Due to the fixed size of the receptive field of the convolution kernel, it is difficult to capture the global features of medical images. Although the self-attention-based Transformer can model long-range dependencies, it has high computational complexity and lacks local inductive bias. Much research has demonstrated that global and local features are crucial for image classification. However, medical images have a lot of noisy, scattered features, intra-class variation, and inter-class similarities. This paper proposes a three-branch hierarchical multi-scale feature fusion network structure termed as HiFuse for medical image classification as a new method. It can fuse the advantages of Transformer and CNN from multi-scale hierarchies without destroying the respective modeling so as to improve the classification accuracy of various medical images. A parallel hierarchy of local and global feature blocks is designed to efficiently extract local features and global representations at various semantic scales, with the flexibility to model at different scales and linear computational complexity relevant to image size. Moreover, an adaptive hierarchical feature fusion block (HFF block) is designed to utilize the features obtained at different hierarchical levels comprehensively. The HFF block contains spatial attention, channel attention, residual inverted MLP, and shortcut to adaptively fuse semantic information between various scale features of each branch. The accuracy of our proposed model on the ISIC2018 dataset is 7.6% higher than baseline, 21.5% on the Covid-19 dataset, and 10.4% on the Kvasir dataset. Compared with other advanced models, the HiFuse model performs the best. Our code is open-source and available from https://github.com/huoxiangzuo/HiFuse.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

huoxiangzuo/HiFuse officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

ClassificationImage ClassificationInductive BiasMedical Image Classificationimage-classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification ISIC 2018 HiFuse_Base F1 75.32 #1 of 2 Archive leaderboard report
Image Classification ISIC 2018 HiFuse_Small F1 72.7 #2 of 2 Archive leaderboard report
Image Classification ISIC2018 HiFuse_Base Accuracy 84.12 #2 of 4 Archive leaderboard report
Image Classification ISIC2018 HiFuse_Small Accuracy 83.59 #3 of 4 Archive leaderboard report
Image Classification ISIC2018 HiFuse_Tiny Accuracy 82.99 #4 of 4 Archive leaderboard report
Image Classification ISIC2018 HiFuse_Tiny F1 72.99 #4 of 4 Archive leaderboard report
Image Classification Kvasir HiFuse_Small Accuracy 85 #1 of 3 Archive leaderboard report
Image Classification Kvasir HiFuse_Small F1 84.96 #1 of 3 Archive leaderboard report
Image Classification Kvasir HiFuse_Tiny Accuracy 84.85 #2 of 3 Archive leaderboard report
Image Classification Kvasir HiFuse_Tiny F1 84.89 #2 of 3 Archive leaderboard report
Image Classification Kvasir HiFuse_Base Accuracy 84.35 #3 of 3 Archive leaderboard report
Image Classification Kvasir HiFuse_Base F1 84.41 #3 of 3 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEConvolutionDense ConnectionsDropoutHierarchical Feature FusionLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections