Kidney and Kidney Tumor Segmentation: MICCAI 2021 Challenge, KiTS 2021, Held in Conjunction with MICCAI 2021, Strasbourg, France, September 27, 2021, Proceedings (Lecture Notes in Computer Science)

Preface
Organization
Contents
Automated Kidney Tumor Segmentation with Convolution and Transformer Network
	1 Introduction
	2 Related Work
		2.1 Medical Image Segmentation
		2.2 Self-attention Mechanism
	3 Methods
		3.1 Network Architecture
		3.2 Loss Function
		3.3 Pre- and post- processing
		3.4 Implementation Details
	4 Results
		4.1 Dataset
		4.2 Metrics
		4.3 Results on KITS21 Training Set
		4.4 Results on KITS21 Test Set
	5 Discussion and Conclusion
	References
Extraction of Kidney Anatomy Based on a 3D U-ResNet with Overlap-Tile Strategy
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
		2.4 Postprocessing
	3 Results
	4 Discussion and Conclusion
	References
Modified nnU-Net for the MICCAI KiTS21 Challenge
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion and Conclusion
	References
2.5D Cascaded Semantic Segmentation for Kidney Tumor Cyst
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion and Conclusion
	References
Automated Machine Learning Algorithm for Kidney, Kidney Tumor, Kidney Cyst Segmentation in Computed Tomography Scans
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Network Architecture
		2.4 Network Training
	3 Results
	4 Discussion and Conclusion
	References
Three Uses of One Neural Network: Automatic Segmentation of Kidney Tumor and Cysts Based on 3D U-Net
	1 Introduction
	2 Methods
		2.1 Network Architecture
		2.2 Segmentation from Low-Resolution CT
		2.3 Fine Segmentation of Kidney
		2.4 Segmentation of Tumor and Cysts
		2.5 Training Protocols
	3 Results
	4 Discussion and Conclusion
	References
Less is More: Contrast Attention Assisted U-Net for Kidney, Tumor and Cyst Segmentations
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Network Architecture
	3 Results
	4 Discussion and Conclusion
	References
A Coarse-to-Fine Framework for the 2021 Kidney and Kidney Tumor Segmentation Challenge
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
		3.1 Metric
		3.2 Results and Discussions
	4 Conclusion
	References
Kidney and Kidney Tumor Segmentation Using a Two-Stage Cascade Framework
	1 Introduction
	2 Methods
		2.1 Kidney-Net
		2.2 Masses-Net
		2.3 Loss Function
	3 Experiment
		3.1 Datasets
		3.2 Pre-processing and Post-processing
		3.3 Training and Implementation Details
		3.4 Metrics
	4 Results and Discussion
	5 Conclusion
	References
Squeeze-and-Excitation Encoder-Decoder Network for Kidney and Kidney Tumor Segmentation in CT Images
	1 Introduction
	2 Method
		2.1 Architecture
		2.2 Squeeze-and-Excitation Module
		2.3 Deep Supervision
		2.4 Loss Function
	3 Experiments
		3.1 Datasets
		3.2 Metrics
		3.3 Pre- and Post-processing
		3.4 Implementation Details
	4 Result
	5 Discussion and Conclusion
	References
A Two-Stage Cascaded Deep Neural Network with Multi-decoding Paths for Kidney Tumor Segmentation
	1 Introduction
	2 Methods
		2.1 Kidney Localization Network
		2.2 Multi-decoding Segmentation Network
		2.3 Global Context Fusion Block
		2.4 Regional Constraint Loss Function
	3 Experimental Results
		3.1 Dataset
		3.2 Implementation Details
		3.3 Results
	4 Conclusion
	References
Mixup Augmentation for Kidney and Kidney Tumor Segmentation
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion
	References
Automatic Segmentation in Abdominal CT Imaging for the KiTS21 Challenge
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion and Conclusion
	References
An Ensemble of 3D U-Net Based Models for Segmentation of Kidney and Masses in CT Scans
	1 Introduction
	2 nnU-Net Determined Details
		2.1 3D U-Net Network Architecture
		2.2 3D U-Net Cascade Network Architecture
		2.3 Preprocessing
		2.4 Training Details
	3 Method
		3.1 Training and Validation Data
		3.2 Pretraining
		3.3 Annotations
		3.4 Regularized Loss
		3.5 Postprocessing
		3.6 Final Submission
	4 Results
		4.1 Single-Stage, High-Resolution 3D U-Net
		4.2 3D U-Net Cascade
		4.3 Model Ensemble
		4.4 Postprocessing
		4.5 Test Set Results
	5 Discussion and Conclusions
	References
Contrast-Enhanced CT Renal Tumor Segmentation
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion and Conclusion
	References
A Cascaded 3D Segmentation Model for Renal Enhanced CT Images
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion and Conclusion
	References
Leveraging Clinical Characteristics for Improved Deep Learning-Based Kidney Tumor Segmentation on CT
	1 Introduction
	2 Materials and Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Baseline 3D U-Net
		2.4 Cognizant Sampling Leveraging Clinical Characteristics
		2.5 Statistical Evaluation
	3 Results
	4 Discussion and Conclusion
	References
A Coarse-to-Fine 3D U-Net Network for Semantic Segmentation of Kidney CT Scans
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Data Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion and Conclusion
	References
3D U-Net Based Semantic Segmentation of Kidneys and Renal Masses on Contrast-Enhanced CT
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Network Architecture
		2.4 Loss Function
		2.5 Optimization Strategy
		2.6 Validation
		2.7 Post-processing
	3 Results
	4 Discussion and Conclusion
	References
Kidney and Kidney Tumor Segmentation Using Spatial and Channel Attention Enhanced U-Net
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Data Augmentations
		2.4 Proposed Method
		2.5 Residual U-Net for Comparison
		2.6 Implementation and Training
		2.7 Inference Procedure
	3 Results
	4 Conclusion
	References
Transfer Learning for KiTS21 Challenge
	1 Introduction
	2 Methods
		2.1 Training and Validation Data
		2.2 Preprocessing
		2.3 Proposed Method
	3 Results
	4 Discussion and Conclusion
	References
Author Index