FMaMIL: Frequency-Driven Mamba Multi-Instance Learning for Weakly Supervised Lesion Segmentation in Medical Images

TL;DR

FMaMIL is a novel weakly supervised lesion segmentation framework that leverages frequency-domain encoding and a two-stage training process to improve accuracy without requiring pixel-level annotations.

Contribution

The paper introduces FMaMIL, combining frequency-domain encoding with a two-stage MIL approach for improved weakly supervised lesion segmentation.

Findings

Outperforms state-of-the-art weakly supervised methods

Effective in handling label noise through self-correction

Validated on multiple histopathology datasets

Abstract

Accurate lesion segmentation in histopathology images is essential for diagnostic interpretation and quantitative analysis, yet it remains challenging due to the limited availability of costly pixel-level annotations. To address this, we propose FMaMIL, a novel two-stage framework for weakly supervised lesion segmentation based solely on image-level labels. In the first stage, a lightweight Mamba-based encoder is introduced to capture long-range dependencies across image patches under the MIL paradigm. To enhance spatial sensitivity and structural awareness, we design a learnable frequency-domain encoding module that supplements spatial-domain features with spectrum-based information. CAMs generated in this stage are used to guide segmentation training. In the second stage, we refine the initial pseudo labels via a CAM-guided soft-label supervision and a self-correction mechanism, enabling robust training even under label noise. Extensive experiments on both public and private histopathology datasets demonstrate that FMaMIL outperforms state-of-the-art weakly supervised methods without relying on pixel-level annotations, validating its effectiveness and potential for digital pathology applications.