scholarly journals Epistemic and Heteroscedastic Uncertainty Estimation in Retinal Blood Vessel Segmentation

2021 ◽  
Vol 7 (3) ◽  
pp. 93-100
Author(s):  
Pedro Costa ◽  
Asim Smailagic ◽  
Jaime Cardoso ◽  
Aurélio Campilho
Keyword(s):  
Medical Image ◽  
State Of The Art ◽  
Ground Truth ◽  
Medical Image Segmentation ◽  
Retinal Blood Vessel ◽  
Medical Specialists ◽  
Current State ◽  
Percentage Points ◽  
Segmentation Methods ◽  
Segmentation Models

Current state-of-the-art medical image segmentation methods require high quality datasets to obtain good performance. However, medical specialists often disagree on diagnosis, hence, datasets contain contradictory annotations. This, in turn, leads to difficulties in the optimization process of Deep Learning models and hinder performance. We propose a method to estimate uncertainty in Convolutional Neural Network (CNN) segmentation models, that makes the training of CNNs more robust to contradictory annotations. In this work, we model two types of uncertainty, heteroscedastic and epistemic, without adding any additional supervisory signal other than the ground-truth segmentation mask. As expected, the uncertainty is higher closer to vessel boundaries, and on top of thinner and less visible vessels where it is more likely for medical specialists to disagree. Therefore, our method is more suitable to learn from datasets created with heterogeneous annotators. We show that there is a correlation between the uncertainty estimated by our method and the disagreement in the segmentation provided by two different medical specialists. Furthermore, by explicitly modeling the uncertainty, the Intersection over Union of the segmentation network improves 5.7 percentage points.

scholarly journals Transformer Based Multi-model Fusion for Medical Image Segmentation

2021 ◽  
Vol 1 (1) ◽  
pp. 50-52
Author(s):  
Bo Dong ◽  
Wenhai Wang ◽  
Jinpeng Li
Keyword(s):  
Image Segmentation ◽  
Medical Image ◽  
State Of The Art ◽  
Medical Image Segmentation ◽  
Model Fusion ◽  
Backbone Network ◽  
Segmentation Task ◽  
Complete Solutions ◽  
Segmentation Models

We present our solutions to the MedAI for all three tasks: polyp segmentation task, instrument segmentation task, and transparency task. We use the same framework to process the two segmentation tasks of polyps and instruments. The key improvement over last year is new state-of-the-art vision architectures, especially transformers which significantly outperform ConvNets for the medical image segmentation tasks. Our solution consists of multiple segmentation models, and each model uses a transformer as the backbone network. we get the best IoU score of 0.915 on the instrument segmentation task and 0.836 on polyp segmentation task after submitting. Meanwhile, we provide complete solutions in https://github.com/dongbo811/MedAI-2021.

scholarly journals Modeling clinical assessor intervariability using deep hypersphere encoder–decoder networks

2019 ◽  
Vol 32 (14) ◽  
pp. 10705-10717 ◽  
Author(s):  
Joost van der Putten ◽  
Fons van der Sommen ◽  
Jeroen de Groof ◽  
Maarten Struyvenberg ◽  
Svitlana Zinger ◽  
...  
Keyword(s):  
State Of The Art ◽  
Ground Truth ◽  
Observer Agreement ◽  
Data Sets ◽  
Observer Variability ◽  
Data Set ◽  
Current State ◽  
Inter Observer Variability ◽  
Proposed Model ◽  
Segmentation Models

AbstractIn medical imaging, a proper gold-standard ground truth as, e.g., annotated segmentations by assessors or experts is lacking or only scarcely available and suffers from large intervariability in those segmentations. Most state-of-the-art segmentation models do not take inter-observer variability into account and are fully deterministic in nature. In this work, we propose hypersphere encoder–decoder networks in combination with dynamic leaky ReLUs, as a new method to explicitly incorporate inter-observer variability into a segmentation model. With this model, we can then generate multiple proposals based on the inter-observer agreement. As a result, the output segmentations of the proposed model can be tuned to typical margins inherent to the ambiguity in the data. For experimental validation, we provide a proof of concept on a toy data set as well as show improved segmentation results on two medical data sets. The proposed method has several advantages over current state-of-the-art segmentation models such as interpretability in the uncertainty of segmentation borders. Experiments with a medical localization problem show that it offers improved biopsy localizations, which are on average 12% closer to the optimal biopsy location.

scholarly journals An Annotation Sparsification Strategy for 3D Medical Image Segmentation via Representative Selection and Self-Training

2020 ◽  
Vol 34 (04) ◽  
pp. 6925-6932 ◽  
Author(s):  
Hao Zheng ◽  
Yizhe Zhang ◽  
Lin Yang ◽  
Chaoli Wang ◽  
Danny Z. Chen
Keyword(s):  
Image Segmentation ◽  
Medical Image ◽  
Piriform Cortex ◽  
Medical Image Segmentation ◽  
Improve Performance ◽  
3D Images ◽  
Data Annotation ◽  
Representative Selection ◽  
Public Datasets ◽  
Segmentation Models

Image segmentation is critical to lots of medical applications. While deep learning (DL) methods continue to improve performance for many medical image segmentation tasks, data annotation is a big bottleneck to DL-based segmentation because (1) DL models tend to need a large amount of labeled data to train, and (2) it is highly time-consuming and label-intensive to voxel-wise label 3D medical images. Significantly reducing annotation effort while attaining good performance of DL segmentation models remains a major challenge. In our preliminary experiments, we observe that, using partially labeled datasets, there is indeed a large performance gap with respect to using fully annotated training datasets. In this paper, we propose a new DL framework for reducing annotation effort and bridging the gap between full annotation and sparse annotation in 3D medical image segmentation. We achieve this by (i) selecting representative slices in 3D images that minimize data redundancy and save annotation effort, and (ii) self-training with pseudo-labels automatically generated from the base-models trained using the selected annotated slices. Extensive experiments using two public datasets (the HVSMR 2016 Challenge dataset and mouse piriform cortex dataset) show that our framework yields competitive segmentation results comparing with state-of-the-art DL methods using less than ∼20% of annotated data.

scholarly journals Synergistic Image and Feature Adaptation: Towards Cross-Modality Domain Adaptation for Medical Image Segmentation

2019 ◽  
Vol 33 ◽  
pp. 865-872 ◽  
Author(s):  
Cheng Chen ◽  
Qi Dou ◽  
Hao Chen ◽  
Jing Qin ◽  
Pheng-Ann Heng
Keyword(s):  
Image Segmentation ◽  
Medical Image ◽  
Domain Adaptation ◽  
State Of The Art ◽  
Medical Image Segmentation ◽  
Target Domain ◽  
Learning Procedure ◽  
The Neural Networks ◽  
Feature Adaptation ◽  
Segmentation Task

This paper presents a novel unsupervised domain adaptation framework, called Synergistic Image and Feature Adaptation (SIFA), to effectively tackle the problem of domain shift. Domain adaptation has become an important and hot topic in recent studies on deep learning, aiming to recover performance degradation when applying the neural networks to new testing domains. Our proposed SIFA is an elegant learning diagram which presents synergistic fusion of adaptations from both image and feature perspectives. In particular, we simultaneously transform the appearance of images across domains and enhance domain-invariance of the extracted features towards the segmentation task. The feature encoder layers are shared by both perspectives to grasp their mutual benefits during the end-to-end learning procedure. Without using any annotation from the target domain, the learning of our unified model is guided by adversarial losses, with multiple discriminators employed from various aspects. We have extensively validated our method with a challenging application of crossmodality medical image segmentation of cardiac structures. Experimental results demonstrate that our SIFA model recovers the degraded performance from 17.2% to 73.0%, and outperforms the state-of-the-art methods by a significant margin.

scholarly journals A Novel Brain Image Segmentation Method Using an Improved 3D U-Net Model

2021 ◽  
Vol 2021 ◽  
pp. 1-10
Author(s):  
Zhuqing Yang
Keyword(s):  
Image Segmentation ◽  
Medical Image ◽  
Three Dimensional ◽  
Research Field ◽  
Medical Image Segmentation ◽  
Tumor Segmentation ◽  
Segmentation Method ◽  
Brain Tumor Segmentation ◽  
Segmentation Methods ◽  
Three Dimensional Segmentation

Medical image segmentation (IS) is a research field in image processing. Deep learning methods are used to automatically segment organs, tissues, or tumor regions in medical images, which can assist doctors in diagnosing diseases. Since most IS models based on convolutional neural network (CNN) are two-dimensional models, they are not suitable for three-dimensional medical imaging. On the contrary, the three-dimensional segmentation model has problems such as complex network structure and large amount of calculation. Therefore, this study introduces the self-excited compressed dilated convolution (SECDC) module on the basis of the 3D U-Net network and proposes an improved 3D U-Net network model. In the SECDC module, the calculation amount of the model can be reduced by 1 × 1 × 1 convolution. Combining normal convolution and cavity convolution with an expansion rate of 2 can dig out the multiview features of the image. At the same time, the 3D squeeze-and-excitation (3D-SE) module can realize automatic learning of the importance of each layer. The experimental results on the BraTS2019 dataset show that the Dice coefficient and other indicators obtained by the model used in this paper indicate that the overall tumor can reach 0.87, the tumor core can reach 0.84, and the most difficult to segment enhanced tumor can reach 0.80. From the evaluation indicators, it can be analyzed that the improved 3D U-Net model used can greatly reduce the amount of data while achieving better segmentation results, and the model has better robustness. This model can meet the clinical needs of brain tumor segmentation methods.

scholarly journals Render U-Net: A Unique Perspective on Render to Explore Accurate Medical Image Segmentation

2020 ◽  
Vol 10 (18) ◽  
pp. 6439
Author(s):  
Chen Li ◽  
Wei Chen ◽  
Yusong Tan
Keyword(s):  
Image Segmentation ◽  
Mortality Rate ◽  
High Mortality ◽  
Medical Image ◽  
Sampling Method ◽  
Medical Image Segmentation ◽  
Natural Image ◽  
High Quality ◽  
Public Datasets ◽  
Segmentation Models

Organ lesions have a high mortality rate, and pose a serious threat to people’s lives. Segmenting organs accurately is helpful for doctors to diagnose. There is a demand for the advanced segmentation model for medical images. However, most segmentation models directly migrated from natural image segmentation models. These models usually ignore the importance of the boundary. To solve this difficulty, in this paper, we provided a unique perspective on rendering to explore accurate medical image segmentation. We adapt a subdivision-based point-sampling method to get high-quality boundaries. In addition, we integrated the attention mechanism and nested U-Net architecture into the proposed network Render U-Net.Render U-Net was evaluated on three public datasets, including LiTS, CHAOS, and DSB. This model obtained the best performance on five medical image segmentation tasks.

Medical segmentation with generative adversarial semi-supervised network

2021 ◽  
Author(s):  
Chuchen Li ◽  
Huafeng Liu
Keyword(s):  
Medical Image ◽  
Large Scale ◽  
Auxiliary Information ◽  
Medical Image Segmentation ◽  
Training Data ◽  
Feature Mapping ◽  
Adversarial Learning ◽  
Segmentation Methods ◽  
Medical Segmentation ◽  
Self Learning

Abstract Recent medical image segmentation methods heavily rely on large-scale training data and high-quality annotations. However, these resources are hard to obtain due to the limitation of medical images and professional annotators. How to utilize limited annotations and maintain the performance is an essential yet challenging problem. In this paper, we try to tackle this problem in a self-learning manner by proposing a Generative Adversarial Semi-supervised Network (GASNet). We use limited annotated images as main supervision signals, and the unlabeled images are manipulated as extra auxiliary information to improve the performance. More specifically, we modulate a segmentation network as a generator to produce pseudo labels for unlabeled images. To make the generator robust, we train an uncertainty discriminator with generative adversarial learning to determine the reliability of the pseudo labels. To further ensure dependability, we apply feature mapping loss to obtain statistic distribution consistency between the generated labels and the real labels. Then the verified pseudo labels are used to optimize the generator in a self-learning manner. We validate the effectiveness of the proposed method on right ventricle dataset, Sunnybrook dataset, STACOM, ISIC dataset, and Kaggle lung dataset. We obtain 0.8402 to 0.9121, 0.8103 to 0.9094, 0.9435 to 0.9724, 0.8635 to 0.886, and 0.9697 to 0.9885 dice coefficient with 1/8 to 1/2 proportion of densely annotated labels, respectively. The improvements are up to 28.6 points higher than the corresponding fully supervised baseline.

Sign in / Sign up

Export Citation Format

Share Document