Robust Brain Magnetic Resonance Image Segmentation for Hydrocephalus Patients: Hard and Soft Attention

  title={Robust Brain Magnetic Resonance Image Segmentation for Hydrocephalus Patients: Hard and Soft Attention},
  author={Xuhua Ren and Jiayu Huo and Kai Xuan and Dongming Wei and Lichi Zhang and Qian Wang},
  journal={2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI)},
  • Xuhua RenJiayu Huo Qian Wang
  • Published 12 January 2020
  • Computer Science
  • 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI)
Brain magnetic resonance (MR) segmentation for hydrocephalus patients is considered as a challenging work. Encoding the variation of the brain anatomical structures from different individuals cannot be easily achieved. The task becomes even more difficult especially when the image data from hydrocephalus patients are considered, which often have large deformations and differ significantly from the normal subjects. Here, we propose a novel strategy with hard and soft attention modules to solve… 

Figures and Tables from this paper

Enhanced brain parcellation via abnormality inpainting for neuroimage-based consciousness evaluation of hydrocephalus patients by lumbar drainage

An innovative inpainting method is designed that can amend the large deformations and lesion erosions in hydrocephalus images, and synthesize the normal brain version without injury, which can effectively support brain parcellation tasks and lay the foundation for the subsequent brain network construction work.

TBI-GAN: An Adversarial Learning Approach for Data Synthesis on Traumatic Brain Segmentation

Experimental results show that the proposed TBI-GAN method can produce synthesized TBI images with high quality and valid label maps, which can greatly improve the 2D and 3D traumatic brain segmentation performance compared with the alternatives.

Deep Learning Achieves Neuroradiologist-Level Performance in Detecting Hydrocephalus Requiring Treatment

Hydcephalus cases requiring treatment can be detected automatically from MRI in a heterogeneous patient population based on quantitative characterization of brain anatomy with performance comparable to that of neuroradiologists.

RCPS: Rectified Contrastive Pseudo Supervision for Semi-Supervised Medical Image Segmentation

This work proposes a novel semi-supervised segmentation method named as Rectified Contrastive Pseudo Supervision (RCPS), which combines a rec- tifled pseudo supervision and voxel-level contrastive learning to improve the effectiveness of semi- super supervised segmentation.

Convolutional Neural Network with Multiscale Fusion and Attention Mechanism for Skin Diseases Assisted Diagnosis

A novel multiscale feature fusion network (MSFA-Net) that can extract feature information at different scales through a multiscales feature fusion structure (MSF) in the network and then calibrate and restore the extracted information to achieve the purpose of melanoma segmentation is proposed.

Attention Aware Deep Learning Model for Wireless Capsule Endoscopy Lesion Classification and Localization

The experiment results show that the proposed lesion aware classification network offers superior classification accuracy thus aggregating semantic and conceptual attention maps using self-attention mechanisms, and helps to improve the model explainability by analyzing the gradients of the attention maps.

MED-TEX: Transfer and Explain Knowledge with Less Data from Pretrained Medical Imaging Models

A small student model is learned with less data by distilling knowledge from a cumbersome pretrained teacher model to address the data-hungry issue and the framework outperforms on the knowledge distillation and model interpretation tasks com-pared to state-of-the-art methods on a fundus dataset.



Multiple Sclerosis Lesion Segmentation from Brain MRI via Fully Convolutional Neural Networks

A fully convolutional neural network (CNN) based method to segment white matter lesions from multi-contrast MR images for multiple sclerosis and significant improvement in segmentation quality over the competing methods is demonstrated.

Transfer Learning for Domain Adaptation in MRI: Application in Brain Lesion Segmentation

This study trained a CNN on legacy MR images of brain and evaluated the performance of the domain-adapted network on the same task with images from a different domain, substantially outperforming a similar network trained on the the same set of examples from scratch.

V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation

This work proposes an approach to 3D image segmentation based on a volumetric, fully convolutional, neural network, trained end-to-end on MRI volumes depicting prostate, and learns to predict segmentation for the whole volume at once.

Nonrigid registration using free-form deformations: application to breast MR images

The results clearly indicate that the proposed nonrigid registration algorithm is much better able to recover the motion and deformation of the breast than rigid or affine registration algorithms.

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

The proposed network extends the previous u-net architecture from Ronneberger et al. by replacing all 2D operations with their 3D counterparts and performs on-the-fly elastic deformations for efficient data augmentation during training.

An Unsupervised Learning Model for Deformable Medical Image Registration

The proposed method uses a spatial transform layer to reconstruct one image from another while imposing smoothness constraints on the registration field, and demonstrates registration accuracy comparable to state-of-the-art 3D image registration, while operating orders of magnitude faster in practice.

Dual Attention Network for Scene Segmentation

New state-of-the-art segmentation performance on three challenging scene segmentation datasets, i.e., Cityscapes, PASCAL Context and COCO Stuff dataset is achieved without using coarse data.

Multiresolution elastic matching

Context Encoding for Semantic Segmentation

The proposed Context Encoding Module significantly improves semantic segmentation results with only marginal extra computation cost over FCN, and can improve the feature representation of relatively shallow networks for the image classification on CIFAR-10 dataset.

Rethinking Atrous Convolution for Semantic Image Segmentation

The proposed `DeepLabv3' system significantly improves over the previous DeepLab versions without DenseCRF post-processing and attains comparable performance with other state-of-art models on the PASCAL VOC 2012 semantic image segmentation benchmark.