# TK-Mamba: Marrying KAN With Mamba for Text-Driven 3D Medical Image Segmentation

Source: [TK-Mamba: Marrying KAN With Mamba for Text-Driven 3D Medical Image Segmentation](<https://www.sota2.com/research/paper/68fa78c91e209131cea7a841>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-11-27T05:01:19.526Z

Authors: Haoyu Yang, Yutong Guan, Meixing Shi, Yuxiang Cai, Jintao Chen, Sun Bing, Wenhui Lei, Mianxin Liu, Xiaoming Shi, Yankai Jiang, Jianwei Yin

Published: 2025-05-24T05:41:55.000Z

## Abstract

3D medical image segmentation is important for clinical diagnosis and treatment but faces challenges from high-dimensional data and complex spatial dependencies. Traditional single-modality networks, such as CNNs and Transformers, are often limited by computational inefficiency and constrained contextual modeling in 3D settings. To alleviate these limitations, we propose TK-Mamba, a multimodal framework that fuses the linear-time Mamba with Kolmogorov-Arnold Networks (KAN) to form an efficient hybrid backbone. Our approach is characterized by two primary technical contributions. Firstly, we introduce the novel 3D-Group-Rational KAN (3D-GR-KAN), which marks the first application of KAN in 3D medical imaging, providing a superior and computationally efficient nonlinear feature transformation crucial for complex volumetric structures. Secondly, we devise a dual-branch text-driven strategy using Pubmedclip's embeddings. This strategy significantly enhances segmentation robustness and accuracy by simultaneously capturing inter-organ semantic relationships to mitigate label inconsistencies and aligning image features with anatomical texts. By combining this advanced backbone and vision-language knowledge, TK-Mamba offers a unified and scalable solution for both multi-organ and tumor segmentation. Experiments on multiple datasets demonstrate that our framework achieves state-of-the-art performance in both organ and tumor segmentation tasks, surpassing existing methods in both accuracy and efficiency. Our code is publicly available at https://github.com/yhy-whu/TK-Mamba

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2505.18525>)

[PDF](<https://arxiv.org/pdf/2505.18525.pdf>)
