Medical Image Computing and Computer Assisted Intervention – MICCAI 2022: 25th International Conference, Singapore, September 18–22, 2022, Proceedings, Part VII (Lecture Notes in Computer Science)

✍ Scribed by Linwei Wang; Qi Dou; P. Thomas Fletcher; Stefanie Speidel; Shuo Li

• Tongue: English
• Leaves: 809
• Category: Library

✦ Table of Contents

Preface
Organization
Contents – Part VII
Image-Guided Interventions and Surgery
Real-Time 3D Reconstruction of Human Vocal Folds via High-Speed Laser-Endoscopy
1 Introduction
2 Method
2.1 Correspondence Matching via Epipolar Constraints
2.2 Surface Reconstruction
3 Results
4 Conclusion
References
Self-supervised Depth Estimation in Laparoscopic Image Using 3D Geometric Consistency
1 Introduction
2 Methodology
2.1 Network Architecture
2.2 Learning 3D Geometric Consistency
2.3 Blind Masking
2.4 Training Loss
3 Experiments
3.1 Dataset
3.2 Evaluation Metrics, Baseline, and Implementation Details
4 Results and Discussion
5 Conclusion
References
USG-Net: Deep Learning-based Ultrasound Scanning-Guide for an Orthopedic Sonographer
1 Introduction
2 Methods
2.1 Dataset Construction
2.2 Anatomical Representation Learning
2.3 USG-Net Architecture
2.4 Loss Function
3 Experiments and Results
4 Discussion and Future Work
5 Conclusions
References
Surgical-VQA: Visual Question Answering in Surgical Scenes Using Transformer
1 Introduction
2 Proposed Method
2.1 Preliminaries
2.2 VisualBERT ResMLP
2.3 VisualBert ResMLP for Classification
2.4 VisualBert ResMLP for Sentence
3 Experiment
3.1 Dataset
3.2 Implementation Details
4 Results
4.1 Ablation Study
5 Discussion and Conclusion
References
DSP-Net: Deeply-Supervised Pseudo-Siamese Network for Dynamic Angiographic Image Matching
1 Introduction
2 Method
2.1 Overview of the Network Architecture
2.2 Pseudo Siamese Attention Dense Block
2.3 Optimization Strategy
2.4 Evaluation Metrics
3 Experiments
3.1 Datasets and Setups
3.2 Comparison with State-of-the-Arts
3.3 Ablation Study
4 Conclusions
References
A Novel Fusion Network for Morphological Analysis of Common Iliac Artery
1 Introduction
2 Method
3 Experiments and Results
4 Conclusion
References
Hand Hygiene Quality Assessment Using Image-to-Image Translation
1 Introduction
2 Methods
2.1 Hand Segmentation and Area Classification
2.2 Translation Between Segmented Hands and Hand Templates
2.3 Loss Function
3 Experiments
3.1 Dataset Construction
3.2 Implementation Details
3.3 Evaluation Metrics
4 Results
5 Conclusion
References
An Optimal Control Problem for Elastic Registration and Force Estimation in Augmented Surgery
1 Introduction
2 Methods
2.1 Hyperelastic Model and Observed Data
2.2 Optimization Problem
2.3 Adjoint Method
3 Results
3.1 Sparse Data Challenge Dataset
3.2 Force Estimation in Robotic Surgery
4 Conclusion
References
PRO-TIP: Phantom for RObust Automatic Ultrasound Calibration by TIP Detection
1 Introduction
2 Approach
2.1 Phantom and Model Preparation
2.2 Calibration Method
3 Experiments and Results
4 Conclusion
References
Multimodal-GuideNet: Gaze-Probe Bidirectional Guidance in Obstetric Ultrasound Scanning
1 Introduction
2 Methods
2.1 Problem Formulation
2.2 Multimodal-GuideNet
3 Experiments
3.1 Data
3.2 Experimental Settings
3.3 Metrics and Baselines
4 Results and Discussion
4.1 Probe Motion Guidance
4.2 Gaze Trajectory Prediction
5 Conclusion
References
USPoint: Self-Supervised Interest Point Detection and Description for Ultrasound-Probe Motion Estimation During Fine-Adjustment Standard Fetal Plane Finding
1 Introduction
2 Methodology
3 Experiments
4 Conclusions
References
Self-supervised 3D Patient Modeling with Multi-modal Attentive Fusion
1 Introduction
2 Methodology
3 Experiments
3.1 2D Keypoint Prediction
3.2 3D Mesh Estimation
3.3 Automated Isocentering with Clinical CT Scans
4 Conclusion
References
SLAM-TKA: Real-time Intra-operative Measurement of Tibial Resection Plane in Conventional Total Knee Arthroplasty
1 Introduction
2 Problem Statement
3 Resection Plane Estimation
3.1 SLAM Formulation for Pin Poses Estimation
3.2 Resection Plane Estimation and the Approach Overview
4 Experiments
4.1 Simulation and Robustness Assessment
4.2 In-vivo Experiments
5 Conclusion
References
Digestive Organ Recognition in Video Capsule Endoscopy Based on Temporal Segmentation Network
1 Introduction
2 Methods
2.1 Study Design (Fig. 1)
2.2 Proposed Organ Recognition Method
3 Results
4 Conclusion
References
Mixed Reality and Deep Learning for External Ventricular Drainage Placement: A Fast and Automatic Workflow for Emergency Treatments
1 Introduction
2 Methods
2.1 Automatic Segmentation
2.2 Automatic Registration
2.3 Registration Accuracy Assessment
3 Results
3.1 Automatic Segmentation
3.2 Registration Accuracy Assessment
4 Conclusions
References
Deep Regression with Spatial-Frequency Feature Coupling and Image Synthesis for Robot-Assisted Endomicroscopy
1 Introduction
2 Methodology
2.1 Spatial-and-Frequency Feature Coupling Network (SFFC-Net)
2.2 Feedback Training (FT)
2.3 Loss Function
3 Experiments and Analysis
4 Conclusion
References
Fast Automatic Liver Tumor Radiofrequency Ablation Planning via Learned Physics Model
1 Introduction
2 Methods
2.1 Physics Emulation via Operator Learning
2.2 Single Electrode Radiofrequency Ablation Planning
3 Experiments and Results
3.1 Data and Pre-processing
3.2 Physics Emulator Evaluation
3.3 Application to Automatic Single Needle RFA Planning
4 Discussion and Conclusion
References
Multi-task Video Enhancement for Dental Interventions
1 Introduction
2 Proposed Method
3 Results and Discussion
References
Outcome and Disease Prediction
Weighted Concordance Index Loss-Based Multimodal Survival Modeling for Radiation Encephalopathy Assessment in Nasopharyngeal Carcinoma Radiotherapy
1 Introduction
2 Materials and Methods
2.1 NPC-REP Dataset Acquisition
2.2 Multimodal Survival Modeling
2.3 WCI Loss Function
3 Experiments
3.1 Implementation Details
3.2 Results and Analysis
4 Conclusions
References
Reducing Positional Variance in Cross-sectional Abdominal CT Slices with Deep Conditional Generative Models
1 Introduction
2 Method
2.1 Technical Background
2.2 C-SliceGen
2.3 Target Slice Selection
3 Experiments and Results
3.1 Dataset
3.2 Implementation Details and Results
4 Discussion and Conclusion
References
Censor-Aware Semi-supervised Learning for Survival Time Prediction from Medical Images
1 Introduction
2 Method
2.1 Model
3 Experiments
3.1 Data Sets
3.2 Experimental Set up
3.3 Results
4 Discussion and Conclusion
References
Prognostic Imaging Biomarker Discovery in Survival Analysis for Idiopathic Pulmonary Fibrosis
1 Introduction
2 Method
2.1 Contrastive Learning of Patch Representations
2.2 Survival Analysis via Clustering ViT
2.3 Novel Prognostic Biomarker Identification
3 Experiment
3.1 Datasets
3.2 Implementation Details
3.3 Experiment Results
4 Discussion
References
Multi-transSP: Multimodal Transformer for Survival Prediction of Nasopharyngeal Carcinoma Patients
1 Introduction
2 Methodology
2.1 Multimodal CNN-Based Encoder
2.2 Transformer-Based Encoder
3 Experiment
3.1 Dataset and Preprocessing
3.2 Experiment Settings
3.3 Comparison with Other Prognosis Prediction Methods
3.4 Ablation Study
4 Conclusion
References
Contrastive Masked Transformers for Forecasting Renal Transplant Function
1 Introduction
2 Related Work
3 Method
3.1 Contrastive Learning for Renal Transplant
3.2 Sequential Model Architecture
3.3 Implementation Details
4 Data
5 Experiments and Analysis
5.1 Ablation Study for Missing Data Strategies
6 Conclusion
References
Assessing the Performance of Automated Prediction and Ranking of Patient Age from Chest X-rays Against Clinicians
1 Introduction
2 Data and Methods
2.1 Dataset
2.2 Age Prediction and Ranking Study
2.3 Age Prediction Models
2.4 Age Conditional Image Generation and Manipulation
3 Results
3.1 Age Prediction Accuracy
3.2 Clinician vs Algorithm Performance Study
3.3 Aging Feature Visualisation
4 Conclusion
References
Transformer Based Multi-task Deep Learning with Intravoxel Incoherent Motion Model Fitting for Microvascular Invasion Prediction of Hepatocellular Carcinoma
1 Introduction
2 Material and Methods
2.1 Study Population, Image Protocol and IVIM Model
2.2 The Proposed Overall Framework
2.3 IVIM Model Parameter Fitting Task
2.4 MVI Classification Task
2.5 Transformer-Based Multi-task Learning
2.6 Implementation
3 Experimental Results
3.1 Evaluation Metrics and Experimental Setup
3.2 Performance Comparison of Multi-task Learning Methods
3.3 Performance Comparison Between the Proposed Method and Single-Task Methods
4 Conclusion
References
Identifying Phenotypic Concepts Discriminating Molecular Breast Cancer Sub-Types
1 Introduction
2 Method
2.1 Classification Network
2.2 Concept Calculation
2.3 Class-Specific Scoring
3 Experiments and Results
3.1 Dataset
3.2 Experimental Setup
3.3 Model Classification Accuracy
3.4 Phenotypic Concepts of Four BC Sub-Types
4 Discussion
References
Fusing Modalities by Multiplexed Graph Neural Networks for Outcome Prediction in Tuberculosis
1 Introduction
2 A Graph Based Multimodal Fusion Framework
3 Experimental Evaluation
3.1 Data and Experimental Setup
3.2 Baselines
3.3 Results
4 Conclusion
References
Deep Multimodal Guidance for Medical Image Classification
1 Introduction
2 Related Work
3 Method
4 Experimental Setup
5 Results and Discussion
6 Conclusion
References
Opportunistic Incidence Prediction of Multiple Chronic Diseases from Abdominal CT Imaging Using Multi-task Learning
1 Introduction
2 Methods
2.1 Dataset
2.2 Multi-planar CT Representation
2.3 Multi-task Learning
2.4 Model Training
3 Results
3.1 Multi-planar CT Representation
3.2 Multi-task Learning
3.3 Test Set Analysis
4 Discussion and Conclusion
References
TMSS: An End-to-End Transformer-Based Multimodal Network for Segmentation and Survival Prediction
1 Introduction
2 Proposed Method
3 Experimental Setup
3.1 Dataset Description
3.2 Data Preprocessing
3.3 Implementation Details
4 Experimental Results
5 Discussion
6 Conclusion and Future Work
References
Surgical Data Science
Bayesian Dense Inverse Searching Algorithm for Real-Time Stereo Matching in Minimally Invasive Surgery
1 Introduction
2 Methodology
2.1 Multiscale DIS
2.2 The Bayesian Patch-Wise Posterior Probability
2.3 The Prior Spatial Gaussian Probability
3 Results and Discussion
3.1 Quantitative Comparisons on the Synthetic Data Set
3.2 Qualitative Comparisons on the In-vivo Dataset
3.3 Processing Rate Comparison
4 Conclusion
References
Conditional Generative Data Augmentation for Clinical Audio Datasets
1 Introduction
2 Materials and Method
2.1 Novel Surgical Audio Dataset
2.2 Data Preprocessing and Baseline Augmentations
2.3 Conditional Generative Data Augmentation Method
2.4 Classification Model
3 Results and Evaluation
4 Discussion
5 Conclusion
References
Rethinking Surgical Instrument Segmentation: A Background Image Can Be All You Need
1 Introduction
2 Proposed Method
2.1 Preliminaries
2.2 Synthesizing Surgical Scenes from a Single Background
3 Experiments
3.1 Datasets
3.2 Implementation Details
3.3 Results and Discussion
3.4 Ablation Studies
4 Conclusion
References
Free Lunch for Surgical Video Understanding by Distilling Self-supervisions
1 Introduction
2 Methodology
2.1 Contrastive Learning on Surgical Videos
2.2 Semantic-Preserving Training of Free Models
2.3 Distilled Self-supervised Learning
3 Experiments
3.1 Datasets
3.2 Implementation Details
3.3 Comparisons with the State-of-the-Art Methods
3.4 Ablation Studies
4 Conclusion
References
Rethinking Surgical Captioning: End-to-End Window-Based MLP Transformer Using Patches
1 Introduction
2 Methodology
2.1 Preliminaries
2.2 Our Model
3 Experiments
3.1 Dataset Description
3.2 Implementation Details
4 Results and Evaluation
5 Ablation Study
6 Discussion and Conclusion
References
CaRTS: Causality-Driven Robot Tool Segmentation from Vision and Kinematics Data
1 Introduction
2 Related Work
3 CaRTS: Causality-Driven Robot Tool Segmentation
4 Experiments
5 Conclusion
References
Instrument-tissue Interaction Quintuple Detection in Surgery Videos
1 Introduction
2 Methodology
2.1 Instrument and Tissue Detection Stage
2.2 Quintuple Prediction Stage
3 Experiments
4 Conclusion
References
Surgical Skill Assessment via Video Semantic Aggregation
1 Introduction
2 Methodology
2.1 Feature Extraction Module
2.2 Semantic Grouping Module
2.3 Temporal Context Modeling Module
2.4 Score Prediction Module
2.5 Incorporating Auxiliary Supervision Signals
3 Experiments
3.1 Baseline Comparison
3.2 Assignment Visualization
3.3 Ablation Study
3.4 Improved Performance with Supervision
4 Conclusion
References
Nonlinear Regression of Remaining Surgical Duration via Bayesian LSTM-Based Deep Negative Correlation Learning
1 Introduction
2 Methods
2.1 Network Architecture
2.2 Training Objectives
2.3 Uncertainty Estimation at Inference Stage
2.4 Implementation Details
3 Experiments
4 Conclusion and Future Work
References
Neural Rendering for Stereo 3D Reconstruction of Deformable Tissues in Robotic Surgery
1 Introduction
2 Method
2.1 Overview of the Neural Rendering-Based Framework
2.2 Deformable Surgical Scene Representations
2.3 Tool Mask-Guided Ray Casting
2.4 Stereo Depth-Cueing Ray Marching
2.5 Optimization for Deformable Radiance Fields
3 Experiments
3.1 Dataset and Evaluation Metrics
3.2 Implementation Details
3.3 Qualitative and Quantitative Results
4 Conclusion
References
Towards Holistic Surgical Scene Understanding
1 Introduction
2 Data and Ground-Truth Annotations
3 Method
3.1 TAPIR
4 Experiments
4.1 Video Feature Extractor Method Comparison
4.2 Results and Discussion
5 Conclusions
References
Multi-modal Unsupervised Pre-training for Surgical Operating Room Workflow Analysis
1 Introduction
2 Related Work
3 Method
3.1 Fusion of Intensity and Depth Maps
3.2 Estimating Code q
4 Experiments
4.1 Datasets
4.2 Surgical Activity Recognition
4.3 Semantic Segmentation
4.4 Remarks
5 Conclusion
References
Deep Laparoscopic Stereo Matching with Transformers
1 Introduction
2 Method
2.1 The HybridStereoNet
2.2 Analyzing Transformer in Laparoscopic Stereo
3 Experiments
3.1 Datasets
3.2 Implementation
3.3 Results
4 Conclusion
References
4D-OR: Semantic Scene Graphs for OR Domain Modeling
1 Introduction
2 Methodology
2.1 Semantic Scene Graphs
2.2 4D-OR Dataset
2.3 Scene Graph Generation
2.4 Use-Case: Clinical Role Prediction
3 Experiments
4 Results and Discussion
5 Conclusion
References
AutoLaparo: A New Dataset of Integrated Multi-tasks for Image-guided Surgical Automation in Laparoscopic Hysterectomy-6pt
1 Introduction
2 Dataset Design and Multi-tasks Formulation
2.1 Dataset Collection
2.2 Task Formulation
2.3 Dataset Annotation and Statistics
2.4 Dataset Analysis and Insights
3 Experiments and Benchmarking Methods
3.1 Workflow Recognition
3.2 Laparoscope Motion Prediction
3.3 Instrument and Key Anatomy Segmentation
4 Conclusion and Future Work
References
Retrieval of Surgical Phase Transitions Using Reinforcement Learning
1 Introduction
2 Methods
2.1 Transition Retrieval Network (TRN)
2.2 Training Details
3 Experiment Setup and Dataset Description
4 Results and Discussion
5 Conclusion
References
SGT: Scene Graph-Guided Transformer for Surgical Report Generation
1 Introduction
2 Methodology
2.1 Overview of the Proposed Framework
2.2 Relation Driven Attention
2.3 Graph Induced Attention
2.4 Caption Generation
3 Experiment Results and Analysis
3.1 Dataset
3.2 Experimental Settings
3.3 Experimental Results
4 Conclusion
References
CLTS-GAN: Color-Lighting-Texture-Specular Reflection Augmentation for Colonoscopy
1 Introduction
2 Related Works
3 Data
4 Methods
5 Results and Discussion
References
Adaptation of Surgical Activity Recognition Models Across Operating Rooms
1 Introduction
2 Related Works
3 Method
3.1 Unsupervised Domain Adaptation (UDA)
3.2 Training on Untrimmed Surgical Videos
3.3 Extension to Semi-supervised Domain Adaptation (SSDA)
3.4 The Importance of Pretraining
4 Experiments
5 Conclusion
References
Video-Based Surgical Skills Assessment Using Long Term Tool Tracking
1 Background and Introduction
2 Materials and Methods
2.1 Dataset Description
2.2 Tracking Algorithm
2.3 Feature Based Skill Assessment
2.4 Learning Based Skill Assessment
3 Results
3.1 Tracking Model Performance
3.2 Skill Assessment on Cholec80
4 Discussion
References
Surgical Scene Segmentation Using Semantic Image Synthesis with a Virtual Surgery Environment
1 Introduction
2 Background
3 Data Generation
4 Experimental Results
5 Conclusion
References
Surgical Planning and Simulation
Deep Learning-Based Facial Appearance Simulation Driven by Surgically Planned Craniomaxillofacial Bony Movement
1 Introduction
2 Method
2.1 Point-Wise Feature Extraction
2.2 Attentive Correspondence Assisted Movement Transformation
3 Experiments and Results
3.1 Dataset
3.2 Implementation and Evaluation Methods
3.3 Results
3.4 Ablation Studies
4 Discussions and Conclusions
References
Deep Learning-Based Head and Neck Radiotherapy Planning Dose Prediction via Beam-Wise Dose Decomposition
1 Introduction
2 Method
2.1 Global Coarse Dose Prediction
2.2 Beam-Wise Dose Prediction
2.3 Training Objective
3 Experiment
3.1 Comparison with State-of-the-Art Methods
3.2 Ablation Study
4 Conclusion
References
Ideal Midsagittal Plane Detection Using Deep Hough Plane Network for Brain Surgical Planning
1 Introduction
2 Methodology
2.1 DHT and IDHT
2.2 Sparse DHT Strategy
2.3 Hough Pyramid Attention Network (HPAN)
2.4 Dual Space Supervision (DSS)
3 Experiments
3.1 Dataset and Implementation Details
3.2 Quantitative and Qualitative Evaluation
4 Conclusion
References
Greedy Optimization of Electrode Arrangement for Epiretinal Prostheses
1 Introduction
2 Related Work
3 Methods
3.1 Phosphene Model
3.2 Dictionary Selection
4 Results
4.1 Visual Subfield Coverage
4.2 Electrode Arrangement
4.3 Comparison with Argus II
5 Conclusion
References
Stereo Depth Estimation via Self-supervised Contrastive Representation Learning
1 Introduction
2 Methods
3 Experimental Results and Discussion
4 Conclusion
References
Deep Geometric Supervision Improves Spatial Generalization in Orthopedic Surgery Planning
1 Introduction
2 Materials and Methods
2.1 Automatic Approach to Orthopedic Surgical Planning
2.2 Deep Geometric Supervision (DGS)
2.3 Model Variants
2.4 Datasets and Training Protocol
3 Results
4 Discussion
References
On Surgical Planning of Percutaneous Nephrolithotomy with Patient-Specific CTRs
1 Introduction
1.1 Concentric-Tube Robots (CTRs)
1.2 Related Work
1.3 Contribution of the Paper
2 Surgical Planning
3 Patient-Specific Design of CTRs
3.1 Problem Formulation
4 Study Cases
5 Results
6 Conclusion
References
Machine Learning – Domain Adaptation and Generalization
Low-Resource Adversarial Domain Adaptation for Cross-modality Nucleus Detection
1 Introduction
2 Methodology
2.1 Target Task-Augmented Bidirectional Adversarial Learning
2.2 Stochastic Data Augmentation
2.3 Implementation Details
3 Experiments
4 Conclusion
References
Domain Specific Convolution and High Frequency Reconstruction Based Unsupervised Domain Adaptation for Medical Image Segmentation
1 Introduction
2 Method
2.1 Problem Definition and Method Overview
2.2 DSC Module
2.3 Encoder-decoder Backbone
2.4 HFR and Segmentation
2.5 Training and Inference
3 Experiments and Results
4 Conclusion
References
Unsupervised Cross-disease Domain Adaptation by Lesion Scale Matching
1 Introduction
2 Methodology
2.1 Preliminary: The Margin Disparity Discrepancy Method
2.2 Latent Space Search for Bounding Box Size
2.3 Monte Carlo Expectation Maximization
3 Experiments
3.1 Experimental Settings
3.2 Experimental Results
4 Conclusion and Future Work
References
Adversarial Consistency for Single Domain Generalization in Medical Image Segmentation
1 Introduction
2 Related Works
3 Proposed Method
3.1 Adversarial Domain Synthesizer
3.2 MI Regularization
3.3 Model Optimization
4 Experiments
4.1 Training Configuration
4.2 Data Prepossessing and Evaluation Metrics
4.3 Experimental Results and Empirical Analysis
4.4 Ablation Study
5 Conclusion
References
Delving into Local Features for Open-Set Domain Adaptation in Fundus Image Analysis
1 Introduction
2 Proposed Methods
2.1 Problem Formulation
2.2 Region-Level Clustering and Adaptation
2.3 Optimization and Inference
3 Experiments and Results
3.1 Dataset Setup and Implementation Details
3.2 Comparison with State-of-the-Arts
3.3 Ablation Studies
4 Conclusion
References
Estimating Model Performance Under Domain Shifts with Class-Specific Confidence Scores
1 Introduction
2 Method
3 Experiments
4 Conclusion
References
vMFNet: Compositionality Meets Domain-Generalised Segmentation
1 Introduction
2 Related Work
3 Proposed Method
3.1 Learning Compositional Components
3.2 Composing Components for Reconstruction
3.3 Performing Downstream Task
3.4 Learning Objective
4 Experiments
4.1 Datasets and Baseline Models
4.2 Implementation Details
4.3 Semi-supervised Domain Generalisation
4.4 Visualisation of Compositionality
4.5 Test-Time Domain Generalisation
5 Conclusion
References
Domain Adaptive Nuclei Instance Segmentation and Classification via Category-Aware Feature Alignment and Pseudo-Labelling
1 Introduction
2 Methods
2.1 Overview
2.2 Category-Aware Feature Alignment
2.3 Nuclei-Level Prototype Pseudo-Labelling
3 Experiments
3.1 Datasets and Evaluation Metrics
3.2 Implementation Details
3.3 Comparison Experiments
3.4 Ablation Studies
4 Conclusion
References
Learn to Ignore: Domain Adaptation for Multi-site MRI Analysis
1 Introduction
1.1 Related Work
1.2 Problem Statement
2 Method
2.1 Loss Functions
3 Experiments
4 Results and Discussion
5 Conclusion
References
Enhancing Model Generalization for Substantia Nigra Segmentation Using a Test-time Normalization-Based Method-6pt
1 Introduction
2 Methods
2.1 Test-time Normalization
2.2 Prior-atlas-based Likelihood Estimation
3 Experiments
3.1 Datasets
3.2 Implementation Details
4 Results
4.1 Quantitative Evaluation
4.2 Qualitative Evaluation
5 Discussion
6 Conclusions
References
Attention-Enhanced Disentangled Representation Learning for Unsupervised Domain Adaptation in Cardiac Segmentation
1 Introduction
2 Methodology
2.1 Overall Framework
2.2 Alignment of Imaging Characteristics
2.3 Channel-wise Disentanglement
2.4 Attention Bias for Adversarial Learning
2.5 Implementation Details
3 Experiment Results and Analysis
3.1 Dataset and Evaluation Metrics
3.2 Effectiveness of ADR in Cardiac Segmentation
3.3 Ablation Study
4 Conclusions
References
Histogram-Based Unsupervised Domain Adaptation for Medical Image Classification*-10pt
1 Introduction
2 Literature Review
3 Methods
3.1 Overview
3.2 Histogram Layer
3.3 Histogram Discriminator
3.4 Graymap GAN
3.5 Gamma-Adjustment GAN
4 Experiments
4.1 Datasets
4.2 Training
5 Results, Discussion, and Conclusions
References
Multi-institutional Investigation of Model Generalizability for Virtual Contrast-Enhanced MRI Synthesis
1 Introduction
2 Materials and Methods
2.1 Data Description
2.2 Deep Learning Network
2.3 Study Design
3 Results and Discussion
4 Conclusion
References
Author Index