Dataset Information

SA-Net: A scale-attention network for medical image segmentation.

ABSTRACT: Semantic segmentation of medical images provides an important cornerstone for subsequent tasks of image analysis and understanding. With rapid advancements in deep learning methods, conventional U-Net segmentation networks have been applied in many fields. Based on exploratory experiments, features at multiple scales have been found to be of great importance for the segmentation of medical images. In this paper, we propose a scale-attention deep learning network (SA-Net), which extracts features of different scales in a residual module and uses an attention module to enforce the scale-attention capability. SA-Net can better learn the multi-scale features and achieve more accurate segmentation for different medical image. In addition, this work validates the proposed method across multiple datasets. The experiment results show SA-Net achieves excellent performances in the applications of vessel detection in retinal images, lung segmentation, artery/vein(A/V) classification in retinal images and blastocyst segmentation. To facilitate SA-Net utilization by the scientific community, the code implementation will be made publicly available.

SUBMITTER: Hu J

PROVIDER: S-EPMC8046243 | biostudies-literature | 2021

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

SA-Net: A scale-attention network for medical image segmentation.

Hu Jingfei J Wang Hua H Wang Jie J Wang Yunqi Y He Fang F Zhang Jicong J

PloS one 20210414 4

Semantic segmentation of medical images provides an important cornerstone for subsequent tasks of image analysis and understanding. With rapid advancements in deep learning methods, conventional U-Net segmentation networks have been applied in many fields. Based on exploratory experiments, features at multiple scales have been found to be of great importance for the segmentation of medical images. In this paper, we propose a scale-attention deep learning network (SA-Net), which extracts features ...[more]

PMID: 33852577

Similar Datasets

Project description:BackgroundThe precise identification of the position and form of a tumor mass can improve early diagnosis and treatment. However, due to the complicated tumor categories and varying sizes and forms, the segregation of brain gliomas and their internal sub-regions is still very challenging. This study sought to design a new deep-learning network based on three-dimensional (3D) U-Net to address its shortcomings in brain tumor segmentation (BraTS) tasks.MethodsWe developed a 3D dilated multi-scale residual attention U-Net (DMRA-U-Net) model for magnetic resonance imaging (MRI) BraTS. It used dilated convolution residual (DCR) modules to better process shallow features, multi-scale convolution residual (MCR) modules in the bottom encoding path to create richer and more comprehensive feature expression while reducing overall information loss or blurring, and a channel attention (CA) module between the encoding and decoding paths to address the problem of retrieving and preserving important features during the processing of deep feature maps.ResultsThe BraTS 2018-2021 datasets served as the training and evaluation datasets for this study. Further, the proposed architecture was assessed using metrics such as the dice similarity coefficient (DSC), Hausdorff distance (HD), and sensitivity (Sens). The DMRA U-Net model segments the whole tumor (WT), and the tumor core (TC), and the enhancing tumor (ET) regions of brain tumors. Using the suggested architecture, the DSCs were 0.9012, 0.8867, and 0.8813, the HDs were 28.86, 13.34, and 10.88 mm, and the Sens was 0.9429, 0.9452, and 0.9303 for the WT, TC, and ET regions, respectively. Compared to the traditional 3D U-Net, the DSC of the DMRA U-Net increased by 4.5%, 2.5%, and 0.8%, the HD of the DMRA U-Net decreased by 21.83, 16.42, and 10.00, the Sens of the DMRA U-Net increased by 0.4%, 0.7%, and 1.4% for the WT, TC, and ET regions, respectively. Further, the results of the statistical comparison of the performance indicators revealed that our model performed well generally in the segmentation of the WT, TC, and ET regions.ConclusionsWe developed a promising tumor segmentation model. Our solution is open sourced and is available at: https://github.com/Gold3nk/dmra-unet.

Project description:BackgroundMedical image segmentation is crucial for clinical diagnostics and treatment planning. Recently, hybrid models often neglect the local modeling capabilities of Transformers for medical image segmentation, despite the complementary nature of local information from both convolutional neural networks (CNNs) and transformers. This limitation is particularly problematic in multi-organ segmentation, where organs are closely adhered, and accurate delineation is essential. This study aims to develop a novel method that leverages the strengths of both CNNs and transformers to address the challenges of segmenting multiple closely adhered organs, improving accuracy and robustness in multi-organ segmentation.MethodsIn this study, we present the Not Another Dual Attention block (NNDA-block), a versatile, plug-and-play module that seamlessly integrates channel and spatial attention mechanisms. This block can be incorporated at any stage within the Not Another Dual Attention UNet Transformer (NNDA-UNETR) framework. Our novel approach to spatial attention uniquely combines local modeling, significantly reducing detail and texture loss during the up- and down-sampling processes typical in U-shaped architectures.ResultsEvaluations on the Beyond the Cranial Vault (BTCV) and Automatic Cardiac Diagnosis Challenge benchmark datasets demonstrate our method's effectiveness in balancing model complexity with accuracy. On the BTCV dataset, our model sets new state-of-the-art benchmarks, achieving a Dice similarity coefficient of 84.13%, normalized surface dice of 86.96%, mean average surface distance of 3.46 mm, and an Hausdorff distance at 95% of 17.76 mm. Moreover, it reduces the parameter count by 34.2% compared to leading contemporary methods, all while maintaining low computational costs [measured in floating point operations per second (FLOPs)].ConclusionsNNDA-UNETR offers a robust solution for accurate segmentation in multi-organ tasks, particularly where organ adhesion poses challenges. Its lightweight design also makes it well-suited for deployment in real-world medical environments with limited computational resources.

Dataset Information

SA-Net: A scale-attention network for medical image segmentation.

Publications

SA-Net: A scale-attention network for medical image segmentation.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets