BeSt-LeS: Benchmarking Stroke Lesion Segmentation using Deep Supervision (2023-10-10T00:00:00.000000Z)

TL;DR

This work considers the publicly available dataset ATLAS $v2.0$ to benchmark various end-to-end supervised U-Net style models and conducts the Wilcoxon test for 3D models to correlate the relationship between predicted and actual stroke volume.

Abstract

Brain stroke has become a significant burden on global health and thus we need remedies and prevention strategies to overcome this challenge. For this, the immediate identification of stroke and risk stratification is the primary task for clinicians. To aid expert clinicians, automated segmentation models are crucial. In this work, we consider the publicly available dataset ATLAS $v2.0$ to benchmark various end-to-end supervised U-Net style models. Specifically, we have benchmarked models on both 2D and 3D brain images and evaluated them using standard metrics. We have achieved the highest Dice score of 0.583 on the 2D transformer-based model and 0.504 on the 3D residual U-Net respectively. We have conducted the Wilcoxon test for 3D models to correlate the relationship between predicted and actual stroke volume. For reproducibility, the code and model weights are made publicly available: https://github.com/prantik-pdeb/BeSt-LeS.

Authors

Prantik Deb

1 papers

Lalith Bharadwaj Baru

1 papers

Kamalaker Dadi

1 papers

TL;DR

Abstract

Authors

References49 items

W-Net: A boundary-enhanced segmentation network for stroke lesions

Ambiguous Medical Image Segmentation Using Diffusion Models

MAPPING: Model Average with Post-processing for Stroke Lesion Segmentation

MONAI: An open-source framework for deep learning in healthcare

Attention 3D U-Net with Multiple Skip Connections for Segmentation of Brain Tumor Images

Automatic Segmentation and Quantitative Assessment of Stroke Lesions on MR Images

SAN-Net: Learning generalization to unseen sites for stroke lesion segmentation with self-adaptive normalization

Cross-Attention and Deep Supervision UNet for Lesion Segmentation of Chronic Stroke

AGMR-Net: Attention-guided multiscale recovery framework for stroke segmentation

A large, curated, open-source stroke neuroimaging dataset to improve lesion segmentation algorithms

TransAttUnet: Multi-Level Attention-Guided U-Net With Transformer for Medical Image Segmentation

To the Best of Our Knowledge

U-Net Transformer: Self and Cross Attention for Medical Image Segmentation

Automated Segmentation of Infarct Lesions in T1-Weighted MRI Scans Using Variational Mode Decomposition and Deep Learning

Ability of weakly supervised learning to detect acute ischemic stroke and hemorrhagic infarction lesions with diffusion-weighted imaging.

nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation

MI-UNet: Multi-Inputs UNet Incorporating Brain Parcellation for Stroke Lesion Segmentation From T1-Weighted Magnetic Resonance Images

Global Stroke Statistics 2019

Automatic post-stroke lesion segmentation on MR images using 3D residual convolutional neural network

Brain Tumor Segmentation Based on 3D Residual U-Net

Weakly Supervised Brain Lesion Segmentation via Attentional Representation Learning

Volumetric Attention for 3D Medical Image Segmentation and Detection

Liver Vessels Segmentation Based on 3d Residual U-NET

D-UNet: A Dimension-Fusion U Shape Network for Chronic Stroke Lesion Segmentation

An attempt at beating the 3D U-Net

X-Net: Brain Stroke Lesion Segmentation Based on Depthwise Separable Convolution and Long-range Dependencies

PHiSeg: Capturing Uncertainty in Medical Image Segmentation

Recurrent residual U-Net for medical image segmentation

A Probabilistic U-Net for Segmentation of Ambiguous Images

Attention U-Net: Learning Where to Look for the Pancreas

Road Extraction by Deep Residual U-Net

Attention is All you Need

Deep Learning for Brain MRI Segmentation: State of the Art and Future Directions

Stroke: a global response is needed

3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation

Efficient multi‐scale 3D CNN with fully connected CRF for accurate brain lesion segmentation

Deep Residual Learning for Image Recognition

U-Net: Convolutional Networks for Biomedical Image Segmentation

Neural Machine Translation by Jointly Learning to Align and Translate

Deep Neural Networks Segment Neuronal Membranes in Electron Microscopy Images

SLIC Superpixels Compared to State-of-the-Art Superpixel Methods

Random Walks for Image Segmentation

Magnetic resonance imaging in acute stroke

The Softmax Nonlinearity: Derivation Using Statistical Mechanics and Useful Properties as a Multiterminal Analog Circuit Element

Descriptor : A large , open source dataset of stroke anatomical brain images and manual lesion segmentations

Critical values and probability levels for the wilcoxon rank sum test and the wilcoxon signed rank test

We have also conducted experiments for both 2D and 3D-based models

BeSt-LeS: Benchmarking

we conduct the Wilcoxon test on 3D models and compare the relationship between predicted and actual stroke volume

Field of Study

Journal Information

Name

Volume