MK-UNet: Multi-kernel Lightweight CNN for Medical Image Segmentation

2025-09-22

AI Summary

To address real-time medical image segmentation under resource-constrained settings (e.g., portable healthcare devices), this paper proposes MK-UNet, an ultra-lightweight multi-kernel U-shaped network. MK-UNet innovatively integrates multi-kernel depthwise separable convolutions with channel-, spatial-, and group-wise gated attention mechanisms within a U-shaped encoder-decoder architecture, enabling joint multi-scale feature extraction and saliency enhancement. With only 0.316M parameters and 0.314G FLOPs, it reduces parameter count by up to 333× compared to state-of-the-art models (e.g., TransUNet, UNeXt). Evaluated on six medical imaging benchmarks, MK-UNet achieves up to a 6.7% improvement in Dice score while maintaining exceptional computational efficiency. This work establishes a new paradigm for high-fidelity, edge-deployable medical diagnosis—balancing accuracy, latency, and hardware feasibility.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Hardware-aware ML
• Reasoning under Uncertainty: Graphical Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

In this paper, we introduce MK-UNet, a paradigm shift towards ultra-lightweight, multi-kernel U-shaped CNNs tailored for medical image segmentation. Central to MK-UNet is the multi-kernel depth-wise convolution block (MKDC) we design to adeptly process images through multiple kernels, while capturing complex multi-resolution spatial relationships. MK-UNet also emphasizes the images salient features through sophisticated attention mechanisms, including channel, spatial, and grouped gated attention. Our MK-UNet network, with a modest computational footprint of only 0.316M parameters and 0.314G FLOPs, represents not only a remarkably lightweight, but also significantly improved segmentation solution that provides higher accuracy over state-of-the-art (SOTA) methods across six binary medical imaging benchmarks. Specifically, MK-UNet outperforms TransUNet in DICE score with nearly 333$ imes$ and 123$ imes$ fewer parameters and FLOPs, respectively. Similarly, when compared against UNeXt, MK-UNet exhibits superior segmentation performance, improving the DICE score up to 6.7% margins while operating with 4.7$ imes$ fewer #Params. Our MK-UNet also outperforms other recent lightweight networks, such as MedT, CMUNeXt, EGE-UNet, and Rolling-UNet, with much lower computational resources. This leap in performance, coupled with drastic computational gains, positions MK-UNet as an unparalleled solution for real-time, high-fidelity medical diagnostics in resource-limited settings, such as point-of-care devices. Our implementation is available at https://github.com/SLDGroup/MK-UNet.

Problem

• Designing ultra-lightweight CNNs for medical image segmentation
• Capturing multi-resolution spatial relationships with multiple kernels
• Achieving high accuracy with minimal computational resources

Innovation

• Multi-kernel depth-wise convolution block for spatial relationships
• Sophisticated attention mechanisms to emphasize salient features
• Ultra-lightweight CNN with minimal parameters and FLOPs