LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.

Since Transformers have demonstrated excellent performance in the segmentation of two-dimensional medical images, recent works have also introduced them into 3D medical segmentation tasks. For example, hierarchical transformers like Swin UNETR have reintroduced several prior knowledge of convolution...

Full description

Saved in:
Bibliographic Details
Main Authors: Ming Li, Jingang Ma, Jing Zhao
Format: Article
Language:English
Published: Public Library of Science (PLoS) 2025-01-01
Series:PLoS ONE
Online Access:https://doi.org/10.1371/journal.pone.0329806
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849228260289806336
author Ming Li
Jingang Ma
Jing Zhao
author_facet Ming Li
Jingang Ma
Jing Zhao
author_sort Ming Li
collection DOAJ
description Since Transformers have demonstrated excellent performance in the segmentation of two-dimensional medical images, recent works have also introduced them into 3D medical segmentation tasks. For example, hierarchical transformers like Swin UNETR have reintroduced several prior knowledge of convolutional networks, further enhancing the model's volumetric segmentation ability on three-dimensional medical datasets. The effectiveness of these hybrid architecture methods is largely attributed to the large number of parameters and the large receptive fields of non-local self-attention. We believe that large-kernel volumetric depthwise convolutions can obtain large receptive fields with fewer parameters. In this paper, we propose a lightweight three-dimensional convolutional network, LKDA-Net, for efficient and accurate three-dimensional volumetric segmentation. This network adopts a large-kernel depthwise convolution attention mechanism to simulate the self-attention mechanism of Transformers. Firstly, inspired by the Swin Transformer module, we investigate different-sized large-kernel convolution attention mechanisms to obtain larger global receptive fields, and replace the MLP in the Swin Transformer with the Inverted Bottleneck with Depthwise Convolutional Augmentation to reduce channel redundancy and enhance feature expression and segmentation performance. Secondly, we propose a skip connection fusion module to achieve smooth feature fusion, enabling the decoder to effectively utilize the features of the encoder. Finally, through experimental evaluations on three public datasets, namely Synapse, BTCV and ACDC, LKDA-Net outperforms existing models of various architectures in segmentation performance and has fewer parameters. Code: https://github.com/zouyunkai/LKDA-Net.
format Article
id doaj-art-db9b35e7dac5494eb3b9e9d5def33b2c
institution Kabale University
issn 1932-6203
language English
publishDate 2025-01-01
publisher Public Library of Science (PLoS)
record_format Article
series PLoS ONE
spelling doaj-art-db9b35e7dac5494eb3b9e9d5def33b2c2025-08-23T05:31:52ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01208e032980610.1371/journal.pone.0329806LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.Ming LiJingang MaJing ZhaoSince Transformers have demonstrated excellent performance in the segmentation of two-dimensional medical images, recent works have also introduced them into 3D medical segmentation tasks. For example, hierarchical transformers like Swin UNETR have reintroduced several prior knowledge of convolutional networks, further enhancing the model's volumetric segmentation ability on three-dimensional medical datasets. The effectiveness of these hybrid architecture methods is largely attributed to the large number of parameters and the large receptive fields of non-local self-attention. We believe that large-kernel volumetric depthwise convolutions can obtain large receptive fields with fewer parameters. In this paper, we propose a lightweight three-dimensional convolutional network, LKDA-Net, for efficient and accurate three-dimensional volumetric segmentation. This network adopts a large-kernel depthwise convolution attention mechanism to simulate the self-attention mechanism of Transformers. Firstly, inspired by the Swin Transformer module, we investigate different-sized large-kernel convolution attention mechanisms to obtain larger global receptive fields, and replace the MLP in the Swin Transformer with the Inverted Bottleneck with Depthwise Convolutional Augmentation to reduce channel redundancy and enhance feature expression and segmentation performance. Secondly, we propose a skip connection fusion module to achieve smooth feature fusion, enabling the decoder to effectively utilize the features of the encoder. Finally, through experimental evaluations on three public datasets, namely Synapse, BTCV and ACDC, LKDA-Net outperforms existing models of various architectures in segmentation performance and has fewer parameters. Code: https://github.com/zouyunkai/LKDA-Net.https://doi.org/10.1371/journal.pone.0329806
spellingShingle Ming Li
Jingang Ma
Jing Zhao
LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.
PLoS ONE
title LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.
title_full LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.
title_fullStr LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.
title_full_unstemmed LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.
title_short LKDA-Net: Hierarchical transformer with large Kernel depthwise convolution attention for 3D medical image segmentation.
title_sort lkda net hierarchical transformer with large kernel depthwise convolution attention for 3d medical image segmentation
url https://doi.org/10.1371/journal.pone.0329806
work_keys_str_mv AT mingli lkdanethierarchicaltransformerwithlargekerneldepthwiseconvolutionattentionfor3dmedicalimagesegmentation
AT jingangma lkdanethierarchicaltransformerwithlargekerneldepthwiseconvolutionattentionfor3dmedicalimagesegmentation
AT jingzhao lkdanethierarchicaltransformerwithlargekerneldepthwiseconvolutionattentionfor3dmedicalimagesegmentation