RF-HiT: Rectified Flow Hierarchical Transformer for General Medical Image Segmentation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908984082956288 |
|---|---|
| author | Djouama, Ahmed Marouane Belaala, Abir Sellam, Abdellah Zakaria Bekhouche, Salah Eddine Distante, Cosimo Hadid, Abdenour |
| author_facet | Djouama, Ahmed Marouane Belaala, Abir Sellam, Abdellah Zakaria Bekhouche, Salah Eddine Distante, Cosimo Hadid, Abdenour |
| contents | Accurate medical image segmentation requires both long-range contextual reasoning and precise boundary delineation, a task where existing transformer- and diffusion-based paradigms are frequently bottlenecked by quadratic computational complexity and prohibitive inference latency. We propose RF-HiT, a Rectified Flow Hierarchical Transformer that integrates an hourglass transformer backbone with a multi-scale hierarchical encoder for anatomically guided feature conditioning. Unlike prior diffusion-based approaches, RF-HiT leverages rectified flow with efficient transformer blocks to achieve linear complexity while requiring only a few discretization steps. The model further fuses conditioning features across resolutions via learnable interpolation, enabling effective multi-scale representation with minimal computational overhead. As a result, RF-HiT achieves a strong efficiency-performance trade-off, requiring only 10.14 GFLOPs, 13.6M parameters, and inference in as few as three steps. Despite its compact design, RF-HiT attains 91.27% mean Dice on ACDC and 87.40% on BraTS 2021, achieving performance comparable to or exceeding that of significantly more intensive architectures. This demonstrates its strong potential as a robust, computationally efficient foundation for real-time clinical segmentation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_19570 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | RF-HiT: Rectified Flow Hierarchical Transformer for General Medical Image Segmentation Djouama, Ahmed Marouane Belaala, Abir Sellam, Abdellah Zakaria Bekhouche, Salah Eddine Distante, Cosimo Hadid, Abdenour Computer Vision and Pattern Recognition Accurate medical image segmentation requires both long-range contextual reasoning and precise boundary delineation, a task where existing transformer- and diffusion-based paradigms are frequently bottlenecked by quadratic computational complexity and prohibitive inference latency. We propose RF-HiT, a Rectified Flow Hierarchical Transformer that integrates an hourglass transformer backbone with a multi-scale hierarchical encoder for anatomically guided feature conditioning. Unlike prior diffusion-based approaches, RF-HiT leverages rectified flow with efficient transformer blocks to achieve linear complexity while requiring only a few discretization steps. The model further fuses conditioning features across resolutions via learnable interpolation, enabling effective multi-scale representation with minimal computational overhead. As a result, RF-HiT achieves a strong efficiency-performance trade-off, requiring only 10.14 GFLOPs, 13.6M parameters, and inference in as few as three steps. Despite its compact design, RF-HiT attains 91.27% mean Dice on ACDC and 87.40% on BraTS 2021, achieving performance comparable to or exceeding that of significantly more intensive architectures. This demonstrates its strong potential as a robust, computationally efficient foundation for real-time clinical segmentation. |
| title | RF-HiT: Rectified Flow Hierarchical Transformer for General Medical Image Segmentation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.19570 |