Abstract:Background and objective: Training convolutional neural networks based on large amount of labeled data has made great progress in the field of image segmentation. However, in medical image segmentation tasks, annotating the data is expensive and time-consuming because pixel-level annotation requires experts in the relevant field. Currently, the combination of consistent regularization and pseudo labeling-based semi-supervised methods has shown good performance in image segmentation. However, in the training process, a portion of low-confidence pseudo labels are generated by the model. And the semi-supervised segmentation method still has the problem of distribution bias between labeled and unlabeled data. The objective of this study is to address the challenges of semi-supervised learning and improve the segmentation accuracy of semi-supervised models on medical images. Methods: To address these issues, we propose an Uncertainty-based Region Clipping Algorithm for semi-supervised medical image segmentation, which consists of two main modules. A module is introduced to compute the uncertainty of two sub-networks predictions with diversity using Monte Carlo Dropout, allowing the model to gradually learn from more reliable targets. To retain model diversity, we use different loss functions for different branches and use Non-Maximum Suppression in one of the branches. The other module is proposed to generate new samples by masking the low-confidence pixels in the original image based on uncertainty information. New samples are fed into the model to facilitate the model to generate pseudo labels with high confidence and enlarge the training data distribution. Results: Comprehensive experiments on the combination of two benchmarks ACDC and BraTS2019 show that our proposed model outperforms state-of-the-art methods in terms of Dice, HD95 and ASD. The results reach an average Dice score of 87.86 % and a HD95 score of 4.214 mm on ACDC dataset. For the brain tumor segmentation, the results reach an average Dice score of 84.79 % and a HD score of 10.13 mm. Conclusions: Our proposed method improves the accuracy of semi-supervised medical image segmentation. Extensive experiments on two public medical image datasets including 2D and 3D modalities demonstrate the superiority of our model. The code is available at: https://github.com/QuintinDong/URCA.

Generalized Probabilistic U-Net for medical image segementation

Effect of latent space distribution on the segmentation of images with multiple annotations

3D U$$^$$-Net: A 3D Universal U-Net for Multi-domain Medical Image Segmentation

Towards Generalizable Medical Image Segmentation with Pixel-wise Uncertainty Estimation

Investigating and Improving Latent Density Segmentation Models for Aleatoric Uncertainty Quantification in Medical Imaging

Quantifying U-Net Uncertainty in Multi-Parametric MRI-based Glioma Segmentation by Spherical Image Projection

Probabilistic U-Net with Kendall Shape Spaces for Geometry-Aware Segmentations of Images

Efficient Bayesian Uncertainty Estimation for nnU-Net

Diff-UNet: A Diffusion Embedded Network for Volumetric Segmentation

SBDF-Net: A Versatile Dual-Branch Fusion Network for Medical Image Segmentation

Probabilistic Deep Learning for Instance Segmentation

Unsupervised Model Adaptation for Source-free Segmentation of Medical Images

CP-UNet: Contour-based Probabilistic Model for Medical Ultrasound Images Segmentation

SUPER-Net: Trustworthy Medical Image Segmentation with Uncertainty Propagation in Encoder-Decoder Networks

Distributional Gaussian Processes Layers for Out-of-Distribution Detection

Probabilistic 3D segmentation for aleatoric uncertainty quantification in full 3D medical data

Uncertainty Quantification in Medical Image Segmentation with Multi-decoder U-Net

URCA: Uncertainty-based region clipping algorithm for semi-supervised medical image segmentation

GU-Net: Causal relationship-based generative medical image segmentation model

Mixed-Precision Quantization of U-Net for Medical Image Segmentation.

Hierarchical Uncertainty Estimation for Medical Image Segmentation Networks