This book constitutes the refereed proceedings of the second International Workshop on Cancer Prevention through Early Detection, CaPTion, held in conjunction with the 26th International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2023, in Vancouver, Canada, in October 2023.
The 11 papers presented at CaPTion 2023 were carefully reviewed and selected from 12 submissions. The workshop invites researchers to submit their work in the field of medical image analysis around the central theme of cancer and early cancer detection, progression, inflammation understanding, multimodality data, and computer-aided navigation.
Author(s): Sharib Ali; Fons van der Sommen; Maureen van Eijnatten; Bartłomiej W. Papież; Yueming Jin; Iris Kolenbrander
Series: Lecture Notes in Computer Science
Edition: 1
Publisher: Springer Nature Switzerland
Year: 2023
Language: English
Pages: x; 144
City: Cham
Tags: Medical Imaging and Computer-Assisted Intervention; MICCAI; Computer Science; Computer Imaging, Vision, Pattern Recognition and Graphics; Machine Learning; Computing Milieux; Computer Applications
Preface
Organization
Contents
Classification
A Deep Attention-Multiple Instance Learning Framework to Predict Survival of Soft-Tissue Sarcoma from Whole Slide Images
1 Introduction
2 Methodology
2.1 Tiles Extraction
2.2 Features Extraction
2.3 Deep Attention-Multiple Instance Learning Model for Survival Prediction
3 Experiments and Results
3.1 Dataset Description and Experimental Setups
3.2 Experimental Results
4 Discussion and Conclusion
References
Towards Real-Time Confirmation of Breast Cancer in the OR Using CNN-Based Raman Spectroscopy Classification
1 Introduction
2 Related Work
3 Materials and Methods
3.1 Clinical Data and Setup
3.2 Workflow and Data Acquisition
3.3 Classification Model
4 Results
4.1 Clinical Data and Data Processing
4.2 Classification Performance
5 Conclusion
References
Fully Automated CAD System for Lung Cancer Detection and Classification Using 3D Residual U-Net with multi-Region Proposal Network (mRPN) in CT Images
1 Introduction
1.1 Contribution
1.2 Paper Organization
2 Our Approach
2.1 Pre-processing
2.2 RPN Split-Merge Cascade Network
2.3 3D Residual U-Net Training Strategy and Architecture
2.4 Malignancy Score-Based Approach (MSBA)
3 Implementation
4 Experimental Results
5 Conclusion
References
Image Captioning for Automated Grading and Understanding of Ulcerative Colitis
1 Introduction
2 Related Work
2.1 Deep-Learning Based UC Scoring
2.2 Automated Medical Report Generation
3 Method
3.1 Feature Extraction
3.2 Sequence Processing and Decoding
4 Experiments and Results
4.1 Implementation Details
4.2 Evaluation Results
5 Discussion and Conclusion
References
Detection and Segmentation
Multispectral 3D Masked Autoencoders for Anomaly Detection in Non-Contrast Enhanced Breast MRI
1 Introduction
1.1 Contribution
2 Related Work
3 Dataset
4 Method
5 Results
6 Discussion
References
Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer
1 Introduction
2 Method
3 Experiments
4 Results
5 Discussion and Conclusion
A Appendix
A.1 Estimating the Mutual Information
A.2 Influence of the Hyperparameter
A.3 HCR Features Extraction
A.4 Model Architecture
References
Assessing the Performance of Deep Learning-Based Models for Prostate Cancer Segmentation Using Uncertainty Scores
1 Introduction
2 Related Work
2.1 Deep Learning Segmentation
2.2 Uncertainty Quantification
3 Materials and Methods
3.1 Dataset
3.2 Uncertainty Estimation in Prostate Segmentation
3.3 Proposed Work
4 Results and Discussion
4.1 Quantitative Results
4.2 Qualitative Results
5 Application
6 Conclusion
References
MoSID: Modality-Specific Information Disentanglement from Multi-parametric MRI for Breast Tumor Segmentation
1 Introduction
2 Methodology
2.1 Step 1: Image Synthesis
2.2 Step 2: Modality-Specific Information Disentanglement
2.3 Step 3: Breast Tumor Segmentation
3 Experiments and Results
3.1 Dataset and Implementation Details
3.2 Segmentation Performance Analysis
4 Conclusion
References
Cancer/Early cancer Surveillance
Colonoscopy Coverage Revisited: Identifying Scanning Gaps in Real-Time
1 Introduction
2 Method: Phase I – Visibility Classification
3 Method: Phase II – Gaps with Loss of Coverage
3.1 Scene Descriptors
3.2 Gap Classification
4 Results
4.1 Scene Descriptors
4.2 Gap Classification
4.3 Implementation Details
5 Conclusion
References
ColNav: Real-Time Colon Navigation for Colonoscopy
1 Introduction
2 Related Work
3 Method Overview
3.1 Centerline and Colon Unfolding
3.2 Navigation Compass
3.3 Unfolding Real-Time Dynamic Update
4 Experiments
4.1 Results
5 Conclusion
References
Modeling Barrett's Esophagus Progression Using Geometric Variational Autoencoders
1 Introduction
1.1 Related Work
2 Method
2.1 VAEs and Hyperspherical VAEs
2.2 Generative Hyperspherical Autoencoder Through a New Loss
2.3 Roto-Equivariant VAE and KS-VAE
3 Experiments
3.1 Dataset
3.2 Experimental Setup
4 Results
5 Discussion
6 Conclusion
References
Author Index
