# Wavelet-enhanced boundary adaptation network for liver hemangioma segmentation in non-contrast CT

## TL;DR

This paper presents WLAU-Net, a new method for segmenting liver hemangiomas in non-contrast CT scans, improving accuracy without using contrast agents.

## Contribution

WLAU-Net introduces a transfer learning framework, wavelet transformation module, and local attention mechanism for contrast-free segmentation.

## Key findings

- WLAU-Net achieved a 65.37% Dice score and 96.23% accuracy, outperforming existing methods.
- The wavelet module significantly contributes to performance, with its removal reducing the Dice score by 1.16%.
- The method maintains high diagnostic accuracy comparable to contrast-enhanced techniques while avoiding allergic risks.

## Abstract

Liver hemangioma segmentation in non-contrast CT images faces significant challenges due to the absence of contrast-enhanced features. This paper introduces WLAU-Net, a novel architecture integrating three key innovations for contrast agent free segmentation. First, our transfer learning framework pre-trains the encoder on venous phase CT images to capture discriminative tumor features, then transfers and freezes these learned weights when processing non-contrast phase data, effectively preventing domain shift. Second, we implement a wavelet transformation module using sym4 wavelet decomposition to split images into four frequency subbands (LL, LH, HL, HH). By selectively amplifying horizontal (HL) and vertical (LH) edge coefficients during reconstruction, we enhance tumor boundary delineation while preserving anatomical context. Third, a local attention mechanism with Gaussian-based adaptive weighting dynamically prioritizes low-intensity tumor regions over high-intensity areas, sharpening focus on subtle boundaries. Experimental results demonstrate WLAU-Net’s superiority with a 65.37% Dice score and 96.23% ACC, outperforming state-of-the-art methods including CS-UNet (64.50% Dice, 93.85% ACC) and Swin-UNet (62.34% Dice, 91.15% ACC). Ablation studies reveal critical contributions from each component: enabling all modules (transfer learning, Gaussian attention, and wavelet enhancement) achieves optimal performance, while removing the wavelet module reduces Dice by 1.16% (64.21%) and disabling both Gaussian and wavelet modules decreases ACC by 3.0% (93.24%). Compared to contrast-enhanced methods (92.1% ACC), our approach maintains competitive diagnostic accuracy (96.23% ACC) while eliminating allergic risks, offering a clinically viable alternative for contrast agent sensitive patients.