Fast and Low-Resource Semi-supervised Abdominal Organ Segmentation: MICCAI 2022 Challenge, FLARE 2022, Held in Conjunction with MICCAI 2022, ... (Lecture Notes in Computer Science)

Preface
Organization
Contents
Self-training with Selective Re-training Improves Abdominal Organ Segmentation in CT Image
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results on Validation Set
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Unlabeled Abdominal Multi-organ Image Segmentation Based on Semi-supervised Adversarial Training Strategy
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Metrics
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Qualitative Results on the Validation Dataset
		4.2 Segmentation Efficiency Results on Validation Set
		4.3 Results on Final Testing Set
		4.4 Limitation and Future Work
	5 Conclusion
	References
Abdominal CT Organ Segmentation by Accelerated nnUNet with a Coarse to Fine Strategy*-4pt
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Network Architecture
		2.3 Training
		2.4 Post-processing
		2.5 Acceleration on Resize and Argmax Operation
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Data Augmentation
		3.3 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Quantitative Results on Final Test Set
		4.3 Qualitative Results on Validation
		4.4 Tricks for Improvement
		4.5 Two Normalization Strategies
		4.6 Effects of Sliding Windows
	5 Conclusion
	References
Semi-supervised Detection, Identification and Segmentation for Abdominal Organs
	1 Introduction
	2 Method
		2.1 nnU-Net
		2.2 Semi-supervised Cascaded Organ Detection, Identification and Segmentation
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 DSC Comparisons Between with and Without Unlabeled Images
		4.3 Visualized Examples of Successful and Failed Cases
		4.4 Segmentation Efficiency Analysis
		4.5 Limitations and Future Work
	5 Conclusion
	References
An Efficiency Coarse-to-Fine Segmentation Framework for Abdominal Organs Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results on Validation Set
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Semi-supervised Augmented 3D-CNN for FLARE22 Challenge
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Segmentation Efficiency Results
	5 Conclusion
	References
DLUNet: Semi-supervised Learning Based Dual-Light UNet for Multi-organ Segmentation
	1 Introduction
	2 Method
		2.1 Consistency-based Learning
		2.2 Light UNet
		2.3 Robust Segmentation Loss
		2.4 Preprocessing and Inference
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Ablation of Semi-supervised Learning
		4.2 Comparison of Loss Function
		4.3 Segmentation Efficiency Results
		4.4 Qualitative Results
		4.5 The Performance of Testing Set
		4.6 Limitation and Future Work
	5 Conclusion
	References
Multi-organ Segmentation Based on 2.5D Semi-supervised Learning
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results on Validation Set
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
3D Cross-Pseudo Supervision (3D-CPS): A Semi-supervised nnU-Net Architecture for Abdominal Organ Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Environments and Requirements
		3.3 Training and Inference Protocols
	4 Results and Discussions
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results
		4.3 Results on Test Dataset
	5 Conclusion
	References
Knowledge Distillation from Cross Teaching Teachers for Efficient Semi-supervised Abdominal Organ Segmentation in CT
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
		3.3 Ablation Study
	4 Results and Discussion
		4.1 Ablation Study
		4.2 Quantitative Results on Validation Set
		4.3 Qualitative Results on Validation Set
		4.4 Segmentation Efficiency Results on Validation Set
		4.5 Results on Test Set
		4.6 Limitations and Future Work
	5 Conclusion
	References
Uncertainty-Guided Self-learning Framework for Semi-supervised Multi-organ Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation and Test Set
		4.2 Qualitative Results
		4.3 Segmentation Efficiency Results
		4.4 Limitations and Future Work
	5 Conclusion
	References
A Noisy nnU-Net Student for Semi-supervised Abdominal Organ Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Inference Optimization
		2.4 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Dataset Splits
		3.3 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results for 5-Fold Cross-Validation
		4.2 Quantitative Results on the Validation and Test Set
		4.3 Qualitative Results
	5 Conclusion
	References
CLEF: Contrastive Learning of Equivariant Features in CT Images*-4pt
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Evaluation on the Validation Set
		4.2 Validation Results
		4.3 Results on Final Testing Set
		4.4 Limitation and Future Work
	5 Conclusion
	References
Teacher-Student Semi-supervised Approach for Medical Image Segmentation
	1 Introduction
	2 Method
		2.1 Phase One
		2.2 Phase Two
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Preprocessing
		3.3 Post-processing
		3.4 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results on Validation Set
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Semi-supervised Organ Segmentation with Mask Propagation Refinement and Uncertainty Estimation for Data Generation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
		2.4 Inference Optimization
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
	5 Conclusion
	References
Revisiting nnU-Net for Iterative Pseudo Labeling and Efficient Sliding Window Inference
	1 Introduction
	2 Method
		2.1 Accurate Segmentation vs Efficient Segmentation
		2.2 Iterative Pseudo Labeling by Accurate Segmentation
		2.3 Efficient Sliding Window Inference
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
A Simple Mean-Teacher UNet Model for Efficient Abdominal Organ Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Quantitative Results on Test Set
		4.4 Segmentation Efficiency Results
		4.5 Limitations and Future Work
	5 Conclusion
	References
Cascade Dual-decoders Network for Abdominal Organs Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Pseudo Labeling
		2.3 Proposed Method
		2.4 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results on Validation Set
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Semi-supervised 3D U-Net Learning Based on Meta Pseudo Labels
	1 Introduction
	2 Method
		2.1 Pre-processing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussions
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results
		4.3 Segmentation Efficiency Results
		4.4 Results on Final Test Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Coarse to Fine Automatic Segmentation of Abdominal Multiple Organs
	1 Introduction
	2 Methods
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
		3.3 Resource Consumption
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
	5 Conclusion
	References
MTSegNet: Semi-supervised Abdominal Organ Segmentation in CT
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Visualization
		4.3 Quantitative Results on Test Set
		4.4 Efficiency of the Method
	5 Conclusion
	References
Uncertainty-aware Mean Teacher Framework with Inception and Squeeze-and-Excitation Block for MICCAI FLARE22 Challenge
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results
		4.1 Ablation Study for Attention Modules
		4.2 Ablation Study for Improvement Strategies
		4.3 Experiments on Different Backbones
		4.4 Comparison Experiments with Baselines
		4.5 Segmentation Results of Our Method
	5 Conclusion
	References
Self-pretrained V-Net Based on PCRL for Abdominal Organ Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Results on Final Testing Set
		4.4 Segmentation Efficiency Results on Validation Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Abdominal Multi-organ Segmentation Using CNN and Transformer
	1 Introduction
	2 Method
		2.1 nnU-Net
		2.2 nnFormer ch24https:spsspsdoi.orgsps10.48550spsarxiv.2109.03201
		2.3 Proposed Method
	3 Experiments
	4 Results
		4.1 Quantitative Results on Validation Set
		4.2 Segmentation Efficiency Results on Validation Set
		4.3 Results on Final Testing Set
		4.4 Limitation and Future Work
	5 Conclusion
	References
Combining Self-training and Hybrid Architecture for Semi-supervised Abdominal Organ Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results on Validation Set
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Semi-supervised Multi-organ Segmentation with Cross Supervision Using Siamese Network
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Cross Supervision Using Siamese Network (CSSN)
		2.4 Unlabeled Image Filtering (UIF) Based on Uncertainty
		2.5 Strategies to Improve Inference Speed and Reduce Resource Consumption
		2.6 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Segmentation Efficiency Results on Validation Set
		4.3 Quantitative Results on Test Set
		4.4 Ablation Study: Influence of Different Number of Unlabeled Data
	5 Discussion and Conclusion
		5.1 Limitation and Future Work
	References
Efficient Semi-supervised Multi-organ Segmentation Using Uncertainty Rectified Pyramid Consistency
	1 Introduction
	2 Method
		2.1 Multi-scale Prediction Network with Pyramid Consistency
		2.2 Uncertainty Rectified Pyramid Consistency Loss
		2.3 The Overall Loss Function
		2.4 Preprocessing
		2.5 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Results on Final Testing Set
	5 Conclusion
	References
A Pseudo-labeling Approach to Semi-supervised Organ Segmentation
	1 Introduction
	2 Method
		2.1 Preprocessing
		2.2 Proposed Method
		2.3 Post-processing
	3 Experiments
		3.1 Dataset and Evaluation Measures
		3.2 Implementation Details
	4 Results and Discussion
		4.1 Quantitative Results on Validation Set
		4.2 Qualitative Results on Validation Set
		4.3 Segmentation Efficiency Results on Validation Set
		4.4 Results on Final Testing Set
		4.5 Limitation and Future Work
	5 Conclusion
	References
Author Index