Compete to Win: Enhancing Pseudo Labels for Barely-supervised Medical Image Segmentation

This study investigates barely-supervised medical image segmentation where only few labeled data, i.e., single-digit cases are available. We observe the key limitation of the existing state-of-the-art semi-supervised solution cross pseudo supervision is the unsatisfactory precision of foreground classes, leading to a degenerated result under barely-supervised learning. In this paper, we propose a novel Compete-to-Win method (ComWin) to enhance the pseudo label quality. In contrast to directly using one model's predictions as pseudo labels, our key idea is that high-quality pseudo labels should be generated by comparing multiple confidence maps produced by different networks to select the most confident one (a compete-to-win strategy). To further refine pseudo labels at near-boundary areas, an enhanced version of ComWin, namely, ComWin+, is proposed by integrating a boundary-aware enhancement module. Experiments show that our method can achieve the best performance on three public medical image datasets for cardiac structure segmentation, pancreas segmentation and colon tumor segmentation, respectively. The source code is now available at https://github.com/Huiimin5/comwin.

[1]  Kecheng Zhang,et al.  ShapePU: A New PU Learning Framework Regularized by Global Consistency for Scribble Supervised Cardiac Segmentation , 2022, MICCAI.

[2]  T. Shinozaki,et al.  FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning , 2022, ICLR.

[3]  Xiaojie Huang,et al.  SSA-Net: Spatial self-attention network for COVID-19 pneumonia infection segmentation with semi-supervised few-shot learning , 2022, Medical Image Analysis.

[4]  G. Carneiro,et al.  Translation Consistent Semi-supervised Segmentation for 3D Medical Images , 2022, ArXiv.

[5]  Philippe Weinzaepfel,et al.  Barely-Supervised Learning: Semi-Supervised Learning with very few labeled images , 2021, AAAI.

[6]  Guotai Wang,et al.  Semi-Supervised Medical Image Segmentation via Cross Teaching between CNN and Transformer , 2021, MIDL.

[7]  Jiliu Zhou,et al.  Semi-supervised NPC segmentation with uncertainty and attention guided consistency , 2021, Knowl. Based Syst..

[8]  N. Ukita,et al.  Embryo Grading With Unreliable Labels Due to Chromosome Abnormalities by Regularized PU Learning With Ranking , 2021, IEEE Transactions on Medical Imaging.

[9]  Qiang Li,et al.  Semi-supervised Learning via Improved Teacher-Student Network for Robust 3D Reconstruction of Stereo Endoscopic Image , 2021, ACM Multimedia.

[10]  Raphael Sznitman,et al.  A Positive/Unlabeled Approach for the Segmentation of Medical Sequences using Point-Wise Supervision , 2021, Medical Image Anal..

[11]  Chuyang Ye,et al.  Positive-unlabeled Learning for Cell Detection in Histopathology Images with Incomplete Annotations , 2021, MICCAI.

[12]  Yuhui Yuan,et al.  Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  Qi Tian,et al.  Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation , 2021, ECCV Workshops.

[14]  Rainer Stiefelhagen,et al.  Every Annotation Counts: Multi-label Deep Supervision for Medical Image Segmentation , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  Jialin Peng,et al.  Medical Image Segmentation With Limited Supervision: A Review of Deep Network Models , 2021, IEEE Access.

[16]  Yinan Chen,et al.  Efficient Semi-supervised Gross Target Volume of Nasopharyngeal Carcinoma Segmentation via Uncertainty Rectified Pyramid Consistency , 2020, MICCAI.

[17]  Han Zhang,et al.  PseudoSeg: Designing Pseudo Labels for Semantic Segmentation , 2020, ICLR.

[18]  Wu-Jun Li,et al.  DMNet: Difference Minimization Network for Semi-supervised Segmentation in Medical Images , 2020, MICCAI.

[19]  Kup-Sze Choi,et al.  Local and Global Structure-Aware Entropy Regularized Mean Teacher Model for 3D Left Atrium Segmentation , 2020, MICCAI.

[20]  Guotai Wang,et al.  Semi-supervised Medical Image Segmentation through Dual-task Consistency , 2020, AAAI.

[21]  Xuming He,et al.  Shape-aware Semi-supervised 3D Semantic Segmentation for Medical Images , 2020, MICCAI.

[22]  Yi Wang,et al.  Deeply-Supervised Networks With Threshold Loss for Cancer Detection in Automated Breast Ultrasound , 2020, IEEE Transactions on Medical Imaging.

[23]  Vicente Ordonez,et al.  Curriculum Labeling: Revisiting Pseudo-Labeling for Semi-Supervised Learning , 2020, AAAI Conference on Artificial Intelligence.

[24]  Gadi Wollstein,et al.  Uncertainty Guided Semi-supervised Segmentation of Retinal Layers in OCT Images , 2019, MICCAI.

[25]  Rynson W. H. Lau,et al.  Dual Student: Breaking the Limits of the Teacher in Semi-Supervised Learning , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[26]  Xiaowei Ding,et al.  Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation , 2019, Medical Image Anal..

[27]  Xilin Chen,et al.  Interlaced Sparse Self-Attention for Semantic Segmentation , 2019, ArXiv.

[28]  Chi-Wing Fu,et al.  Uncertainty-aware Self-ensembling Model for Semi-supervised 3D Left Atrium Segmentation , 2019, MICCAI.

[29]  Hao Chen,et al.  Transformation-Consistent Self-Ensembling Model for Semisupervised Medical Image Segmentation , 2019, IEEE Transactions on Neural Networks and Learning Systems.

[30]  Ronald M. Summers,et al.  A large annotated medical image dataset for the development and evaluation of segmentation algorithms , 2019, ArXiv.

[31]  Wei Shen,et al.  Semi-Supervised 3D Abdominal Multi-Organ Segmentation Via Deep Multi-Planar Co-Training , 2018, 2019 IEEE Winter Conference on Applications of Computer Vision (WACV).

[32]  Hao Chen,et al.  Semi-supervised Skin Lesion Segmentation via Transformation Consistent Self-ensembling Model , 2018, BMVC.

[33]  Xin Yang,et al.  Deep Learning Techniques for Automatic MRI Cardiac Multi-Structures Segmentation and Diagnosis: Is the Problem Solved? , 2018, IEEE Transactions on Medical Imaging.

[34]  Bo Wang,et al.  Deep Co-Training for Semi-Supervised Image Recognition , 2018, ECCV.

[35]  Abhinav Gupta,et al.  Non-local Neural Networks , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[36]  Alan L. Yuille,et al.  Recurrent Saliency Transformation Network: Incorporating Multi-stage Visual Cues for Small Organ Segmentation , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[37]  Ben Glocker,et al.  Semi-supervised Learning for Network-Based Cardiac MR Image Segmentation , 2017, MICCAI.

[38]  Lin Yang,et al.  Deep Adversarial Networks for Biomedical Image Segmentation Utilizing Unannotated Images , 2017, MICCAI.

[39]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[40]  Yang Zhang,et al.  Bayesian belief network for positive unlabeled learning with uncertainty , 2017, Pattern Recognit. Lett..

[41]  Bo Du,et al.  Deeply-supervised CNN for prostate segmentation , 2017, 2017 International Joint Conference on Neural Networks (IJCNN).

[42]  Gang Niu,et al.  Positive-Unlabeled Learning with Non-Negative Risk Estimator , 2017, NIPS.

[43]  Harri Valpola,et al.  Weight-averaged consistency targets improve semi-supervised deep learning results , 2017, ArXiv.

[44]  Seyed-Ahmad Ahmadi,et al.  V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation , 2016, 2016 Fourth International Conference on 3D Vision (3DV).

[45]  Tolga Tasdizen,et al.  Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning , 2016, NIPS.

[46]  Gang Niu,et al.  Theoretical Comparisons of Positive-Unlabeled Learning against Positive-Negative Learning , 2016, NIPS.

[47]  Ronald M. Summers,et al.  DeepOrgan: Multi-level Deep Convolutional Networks for Automated Pancreas Segmentation , 2015, MICCAI.

[48]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[49]  Dacheng Tao,et al.  Classification with Noisy Labels by Importance Reweighting , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[50]  권홍우,et al.  Bootstrapping , 2008, Moral Literacy.

[51]  Vikas Sindhwani,et al.  An RKHS for multi-view learning and manifold co-regularization , 2008, ICML '08.

[52]  Yoshua Bengio,et al.  Semi-supervised Learning by Entropy Minimization , 2004, CAP.

[53]  B. Liu,et al.  Learning to Classify Texts Using Positive and Unlabeled Data , 2003, IJCAI.

[54]  Philip S. Yu,et al.  Partially Supervised Classification of Text Documents , 2002, ICML.

[55]  Sanjoy Dasgupta,et al.  PAC Generalization Bounds for Co-training , 2001, NIPS.

[56]  Avrim Blum,et al.  The Bottleneck , 2021, Monopsony Capitalism.

[57]  David Yarowsky,et al.  Unsupervised Word Sense Disambiguation Rivaling Supervised Methods , 1995, ACL.

[58]  Yu-Feng Li,et al.  Class-Imbalanced Semi-Supervised Learning with Adaptive Thresholding , 2022, ICML.

[59]  Stephen Lin,et al.  Swin Transformer: Hierarchical Vision Transformer using Shifted Windows , 2021, 2021 IEEE/CVF International Conference on Computer Vision (ICCV).

[60]  Mikhail Belkin,et al.  A Co-Regularization Approach to Semi-supervised Learning with Multiple Views , 2005 .

[61]  Yoram Singer,et al.  Unsupervised Models for Named Entity Classification , 1999, EMNLP.