HMC-transducer: hierarchical mamba-CNN transducer for robust liver tumor segmentation - Takeaways - MDSpire

HMC-transducer: hierarchical mamba-CNN transducer for robust liver tumor segmentation

  • By

  • Jiyun Zhu

  • Chao Xu

  • Chang Lei

  • Guangji Zhang

  • Sizhe Fang

  • Shaojun Zhang

  • Jiabin Chen

  • Xuguang Wang

  • January 23, 2026

  • 0 min

Share

  • 1

    Accurate liver tumor segmentation from CT scans is essential for clinical diagnosis and treatment planning.

  • 2

    Existing deep learning models struggle with the trade-off between local feature capture and long-range spatial dependency modeling.

  • 3

    The hierarchical mamba-CNN transducer (HMC-transducer) integrates CNN strengths with Mamba's long-range modeling capabilities.

  • 4

    Key innovations include a direction-aware 3D Mamba block and a gated fusion mechanism for adaptive feature combination.

  • 5

    HMC-Transducer achieves state-of-the-art segmentation accuracy and improved generalization and efficiency on public benchmarks.

Original Source(s)

HMC-transducer: hierarchical mamba-CNN transducer for robust liver tumor segmentation

  • by Jiyun Zhu, Chao Xu, Chang Lei, Guangji Zhang, Sizhe Fang, Shaojun Zhang, Jiabin Chen, Xuguang Wang

  • January 23, 2026

Related Content

Agreement between in vivo and ex vivo photon-counting CT measurements of subchondral trabecular bone features in patients with knee osteoarthritis

Mandatory Training Modules Deserve a Harder Look

A JAMA Viewpoint argues that 4 million physician-hours spent on compliance modules each year warrants institutional reassessment.

Anthropometry, sex, and age at diagnosis affect pulmonary blood volume quantification from computed tomography pulmonary angiography in pulmonary hypertension assessment