Skip to content
Log in
Open access

REC-UNet: a 2D U-Net model with residual cross-dimensional attention for liver tumor segmentation

Zhiyuan Wang Lijun Liang Wei Wu Xugang Xi Yi Lu
Aug 2026 · PeerJ Computer Science · Vol 12, pp. e3986 · 0 citations · 28 references

TL;DR

A novel residual “Enhancement-Calibration” U-Net architecture, termed REC-UNet, which achieves high overall segmentation accuracy across diverse lesion sizes and contrast conditions without relying on explicit size-stratified optimization.

Abstract

Liver tumors impose a significant global health burden, underscoring the urgent need for efficient and accurate diagnostic methods. Computer-assisted techniques, particularly deep learning-based segmentation models, have shown considerable promise in this domain. However, they continue to face persistent challenges in liver tumor segmentation, including severe background noise interference, indistinct lesion boundaries, and the difficulty of simultaneously improving segmentation accuracy while maintaining a balanced trade-off between Recall and Precision. To address these issues, this article proposes a novel residual “Enhancement-Calibration” U-Net architecture, termed REC-UNet. The model consists of two task-specific modules: a Residual Enhancement Module (REM) and a Calibration Module (CM). REM leverages residual connections and cross-dimensional attention to enhance tumor feature representation for accurate segmentation, thereby establishing a foundation for balancing Recall and Precision. CM further mitigates noise propagation from shallow to deep feature layers, refining segmentation precision while sustaining high levels of both Recall and Precision. Experiments on the LiTS2017 and MSD_Task08 liver tumor datasets demonstrate that REC-UNet achieves superior performance over mainstream models, with a 4.34% improvement in Dice and a 4.24% improvement in Intersection over Union (IoU) over the second-best model (VM-UNet) on LiTS2017. We further validate the model on an in-house clinical liver tumor Magnetic Resonance Imaging (MRI) dataset, where it attains a Dice score of 87.88% and an IoU of 86.91%, while maintaining a well-balanced trade-off between Recall and Precision. Importantly, REC-UNet achieves high overall segmentation accuracy across diverse lesion sizes and contrast conditions without relying on explicit size-stratified optimization. These results confirm the robust generalizability of REC-UNet and highlight its significant clinical value for computer-assisted liver tumor diagnosis.

DOI Semantic Scholar
Read PDF

Similar papers

Open access Jul 2026

A 3D residual U-Net with attention-driven spatial pyramid pooling for accurate multimodal MRI brain tumor segmentation

Medical imaging plays a crucial role in the accurate detection and localization of brain tumors, which is essential for effective clinical diagnosis and treatment planning. However, conventional segmentation approaches often struggle to capture complex spatial dependencies in volumetric data. To address this limitation, this study proposes an enhanced 3D U-Net architecture for multi-modal MRI-based brain tumor segmentation. The proposed model leverages three-dimensional convolutional operations to effectively capture contextual and spatial information from volumetric inputs. Additionally, an automated preprocessing pipeline, including image resizing, intensity normalization, and data augmentation, is incorporated to improve model robustness and generalization. The performance of the proposed model is evaluated against a conventional U-Net and a ResNet-based segmentation model using standard metrics such as Dice coefficient, accuracy, Intersection-over-Union (IoU), precision, recall, and F1-score. Experimental results demonstrate that the proposed 3D U-Net achieves superior performance, with a Dice coefficient of 0.83 and a Jaccard index of 0.82, outperforming baseline models across all evaluation metrics. Furthermore, the model exhibits improved convergence behavior and reduced overfitting, indicating strong generalization capability. These findings highlight the effectiveness of the proposed approach for volumetric medical image segmentation. Future work will focus on optimizing hyperparameters, enhancing architectural design, and validating the model on larger and more diverse clinical datasets.

Retinderdeep Singh, C. Prabha, Navita Gupta et al. · 0 citations
Conference Aug 2026

BE-Unet: a boundary-enhanced UNet for skin lesion segmentation

Automated skin lesion segmentation is critical for computer-aided diagnosis (CAD) of dermatological diseases. While UNet and its variants are widely adopted, they face challenges in extracting robust contextual features and aggregating discriminative information due to intrinsic intra-class variation, inter-class similarity, and noise interference in medical images. To address these limitations, we propose BE-Unet, a novel Boundary-Enhanced UNet architecture. BE-Unet integrates three key innovations: Large Paralleling Kernel Attention (LPKA) enhances focus on critical features prone to degradation during early encoding and late decoding stages. Dynamic Weighted Group Multi-axis Hadamard Attention (DHA) augments global-local feature perception. Boundary Enhancement Module (BEM) explicitly addresses ambiguous boundaries via dual-task learning, comprising: (1) Segmentation-Boundary Generation (SBG): Simultaneously predicts lesion masks and boundary maps. (2) Feature Enhancement Fusion (FEF): Dynamically fuses multi-scale features using segmentation and boundary guidance. Extensive experiments on ISIC2017 and ISIC2018 datasets demonstrate that BE-Unet significantly outperforms state-of-the-art methods in segmentation accuracy and boundary delineation, with a parameters count limited to 38 KB and Giga-Operations Per Second (GFLOPs) limited to 0.1.

Xiaoman Huang, Zhenkun Lu · 0 citations