ISLES Challenge: U-Shaped Convolution Neural Network with Dilated Convolution for 3D Stroke Lesion Segmentation

Turečková, Alžběta; Rodríguez-Sánchez, Antonio Jose

doi:10.1007/978-3-030-11723-8_32

Cited by 21 publications

(21 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This paper focuses on how to accurately segment 3D lesions on T1 MRI images of ischemic stroke. Many studies previously used the 3D U-net model to capture the 3D semantic contextual information in three directions by the 3D convolution operation to achieve higher precise segmentation [13], [16]. However, due to the significant number of 3D U-net parameters, it is difficult to train the model with insufficient training data.…”

Section: Related Work a Prediction Methodsmentioning

confidence: 99%

A Partitioning-Stacking Prediction Fusion Network Based on an Improved Attention U-Net for Stroke Lesion Segmentation

Hui

Zhang

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Due to the narrow time window for the treatment of acute ischemic stroke, the stroke lesion area in the patient must be identified as quickly and accurately as possible to evaluate the risks and get the most timely and effective treatment. Therefore, there is great clinical significance in the study of automatic identification and segmentation methods for stroke lesions. In this paper, we propose a partitioningstacking prediction fusion (PSPF) method based on an improved attention U-net to solve the problems of 3D-CNN-based networks, including their high computational cost and insufficient training data, and to achieve accurate segmentation of 3D stroke lesions. Our proposed PSPF method includes three steps in the first part. In Step 1, partitioning, we partition the slices obtained in a certain plane direction by slicing a Magnetic Resonance Imaging (MRI) into subsets according to the 2D graph similarity, then use each partitioned subset to perform training and prediction separately. In Step 2, stacking, we stack the 2D slice results of all subsets according to the position order in MRI before slicing and partitioning to form a 3D lesion result. In Step 3, fusion, we use soft voting to fuse the three orthogonal planes' 3D results that were obtained voxel by voxel in Steps 1 and 2. In the second part, we propose an improved attention U-net, which uses the features from three different scales to generate the attention gating coefficients that further improve training efficiency and segmentation accuracy. We implement a 6-fold cross-validation on the Anatomical Tracings of Lesions After Stroke (ATLAS) dataset to validate our method and model using metrics such as Dice Coefficient (DC), F2 score, and Area under Precision-Recall curve (APR). The results show that compared to the existing methods, our proposed method can not only improve the segmentation precision on unbalanced data but also improve the detailed performance of lesion segmentation. Our proposed method and model are generalized and accurate, demonstrating the good potential for clinical routines. The source codes and models in our method have been made publicly available at [available.upon.acceptance].

show abstract

Section: Related Work a Prediction Methodsmentioning

confidence: 99%

A Partitioning-Stacking Prediction Fusion Network Based on an Improved Attention U-Net for Stroke Lesion Segmentation

Hui

Zhang

et al. 2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…In [169], the authors introduced dilated convolution concept in image segmentation. The conventional convolution networks used downsampling to reduce provided image and upsampling to bring the compressed image back to its original shape.…”

Section: Dilated Mechanismmentioning

confidence: 99%

Deep Neural Architectures for Medical Image Semantic Segmentation: Review

et al. 2021

View full text Add to dashboard Cite

Deep learning has an enormous impact on medical image analysis. Many computer-aided diagnostic systems equipped with deep networks are rapidly reducing human intervention in healthcare. Among several applications, medical image semantic segmentation is one of the core areas of active research to delineate the anatomical structures and other regions of interest. It has a significant contribution to healthcare and provides guided interventions, radiotherapy, and improved radiological diagnostics. The underlying article provides a brief overview of deep convolutional neural architecture, the platforms and applications of deep neural networks, metrics used for empirical evaluation, state-of-the-art semantic segmentation architectures based on a foundational convolution concept, and a review of publicly available medical image datasets highlighting four distinct regions of interest. The article also analyzes the existing work and provides open-ended potential research directions in deep medical image semantic segmentation.

show abstract

“…More recent works build upon the U-Net [22], a 2D fully-convolutional network with skip connections and up-convolutions. For example, [16] used a Dual Path Network [5] encoder while [29] leveraged dilated convolutions to inexpensively increase receptive fields. Furthermore, [1] fused the U-net with other highperforming modules, the BConvLSTM [27] and the SENet [10], and [21] introduced X-blocks to the U-Net, leveraging depthwise separable convolutions to reduce computational load.…”

Section: Related Workmentioning

confidence: 99%

Small Lesion Segmentation in Brain MRIs with Subpixel Embedding

Wong¹,

Chen²,

Wu³

et al. 2021

Preprint

View full text Add to dashboard Cite

We present a method to segment MRI scans of the human brain into ischemic stroke lesion and normal tissues. We propose a neural network architecture in the form of a standard encoder-decoder where predictions are guided by a spatial expansion embedding network. Our embedding network learns features that can resolve detailed structures in the brain without the need for high-resolution training images, which are often unavailable and expensive to acquire. Alternatively, the encoderdecoder learns global structures by means of striding and max pooling. Our embedding network complements the encoder-decoder architecture by guiding the decoder with fine-grained details lost to spatial downsampling during the encoder stage. Unlike previous works, our decoder outputs at 2× the input resolution, where a single pixel in the input resolution is predicted by four neighboring subpixels in our output. To obtain the output at the original scale, we propose a learnable downsampler (as opposed to hand-crafted ones e.g. bilinear) that combines subpixel predictions. Our approach improves the baseline architecture by ≈ 11.7% and achieves the state of the art on the ATLAS public benchmark dataset with a smaller memory footprint and faster runtime than the best competing method. Our source code has been made available at: https://github.com/alexklwong/subpixel-embedding-segmentation.

show abstract

ISLES Challenge: U-Shaped Convolution Neural Network with Dilated Convolution for 3D Stroke Lesion Segmentation

Cited by 21 publications

References 14 publications

A Partitioning-Stacking Prediction Fusion Network Based on an Improved Attention U-Net for Stroke Lesion Segmentation

A Partitioning-Stacking Prediction Fusion Network Based on an Improved Attention U-Net for Stroke Lesion Segmentation

Deep Neural Architectures for Medical Image Semantic Segmentation: Review

Small Lesion Segmentation in Brain MRIs with Subpixel Embedding

Contact Info

Product

Resources

About