Rethinking the Nested U-Net Approach: Enhancing Biomarker Segmentation with Attention Mechanisms and Multiscale Feature Fusion

2025-04-08

AI Summary

To address two key challenges in medical image segmentation—insufficient feature extraction due to morphological and staining variability, and failure of multi-scale feature propagation under limited training samples—this paper proposes a nested U-Net architecture. The method introduces a novel dual-path attention mechanism (combining channel and spatial attention) alongside a multi-scale feature reuse strategy to enhance encoder feature utilization and decoder-based fine-grained reconstruction. Within the U-Net++ framework, we integrate CBAM modules, cross-level feature pyramid fusion, and deep supervision losses. Evaluated on four biomedical datasets, our approach achieves average Dice scores 2.3%–4.7% higher than state-of-the-art methods. Ablation studies quantitatively validate the contribution of each component. The source code is publicly available.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Identifying biomarkers in medical images is vital for a wide range of biotech applications. However, recent Transformer and CNN based methods often struggle with variations in morphology and staining, which limits their feature extraction capabilities. In medical image segmentation, where data samples are often limited, state-of-the-art (SOTA) methods improve accuracy by using pre-trained encoders, while end-to-end approaches typically fall short due to difficulties in transferring multiscale features effectively between encoders and decoders. To handle these challenges, we introduce a nested UNet architecture that captures both local and global context through Multiscale Feature Fusion and Attention Mechanisms. This design improves feature integration from encoders, highlights key channels and regions, and restores spatial details to enhance segmentation performance. Our method surpasses SOTA approaches, as evidenced by experiments across four datasets and detailed ablation studies. Code: https://github.com/saadwazir/ReN-UNet

Problem

• Enhancing biomarker segmentation in medical images
• Addressing limitations in feature extraction from morphology variations
• Improving multiscale feature transfer in encoder-decoder architectures

Innovation

• Nested UNet with multiscale feature fusion
• Attention mechanisms for key regions
• Enhanced encoder-decoder feature integration