Clinical Image-Based Procedures, Fairness of AI in Medical Imaging, and Ethical and Philosophical Issues in Medical Imaging

Name: Clinical Image-Based Procedures, Fairness of AI in Medical Imaging, and Ethical and Philosophical Issues in Medical Imaging | 12th International Workshop, CLIP 2023 1st International Workshop, FAIMI 2023 and 2nd International Workshop, EPIMI 2023 Vancouver, BC, Canada, October 8 and October 12, 2023 Proceedings
Brand: Springer
Price: 106.99 EUR
Availability: OnlineOnly

12th International Workshop, CLIP 2023 1st International Workshop, FAIMI 2023 and 2nd International Workshop, EPIMI 2023 Vancouver, BC, Canada, October 8 and October 12, 2023 Proceedings

Stefan Wesarg Esther Puyol Antón John S. H. Baxter Marius Erdt Klaus Drechsler Cristina Oyarzun Laura Moti Freiman Yufei Chen Islem Rekik Roy Eagleson Aasa Feragen Andrew P. King Veronika Cheplygina Melani Ganz-Benjaminsen Enzo Ferrante Ben Glocker Daniel Moyer Eikel Petersen(Herausgeber*in)

Springer (Verlag)

Erschienen am 9. Oktober 2023

XXII, 314 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-031-45249-9 (ISBN)

106,99 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
Additional Editors
CLIP Preface
CLIP Organization
FAIMI Preface
FAIMI Organization
EPIMI Preface
EPIMI Organization
Contents
CLIP
Automated Hand Joint Classification of Psoriatic Arthritis Patients Using Routinely Acquired Near Infrared Fluorescence Optical Imaging
1 Introduction
2 Background
3 Method
4 Results
5 Discussion and Future Work
References
Automatic Neurocranial Landmarks Detection from Visible Facial Landmarks Leveraging 3D Head Priors
1 Introduction
2 Methods
2.1 Datasets and Preprocessing
2.2 Models Training and Evaluation
3 Experimental Results
3.1 Neurocranial Landmark Coordinates Prediction
3.2 3DMM Validation
3.3 Ablation Study
4 Discussion and Conclusions
References
Subject-Specific Modelling of Knee Joint Motion for Routine Pre-operative Planning
1 Introduction
2 Method
2.1 Contact Surface Model of PF and TF Joint
2.2 Computation of Knee Flexion Angle
2.3 Matching Tibia and Patella Poses
3 Experiments and Discussions
3.1 Evaluation of Generated Patella and Tibia Poses
3.2 Evaluation of Tibia and Patella Pose Matching
4 Conclusion
References
Towards Fine-Grained Polyp Segmentation and Classification
1 Introduction
2 Method
2.1 Swin Transformer Encoder
2.2 Multi-Scale Feature Enhancement
2.3 Patch-Expanding Decoder
2.4 Upsample Head
2.5 Loss Function
3 PolypSegm-ASH Dataset
4 Results
4.1 Experiments on PolypSegm-ASH
4.2 Experiments on Binary Polyp Segmentation
4.3 Ablation Study. Effect of Up-Samples Before Predictions
5 Conclusion
References
Automated Orientation and Registration of Cone-Beam Computed Tomography Scans
1 Introduction
2 Materials
3 Proposed Method
3.1 Automated Standardized Orientation (ASO)
3.2 Automated Registration (AReg)
3.3 Evaluation Metrics
3.4 Implementation
4 Results
4.1 Orientation
4.2 Registration
5 Discussion
6 Conclusion
A Appendix
References
Deep Learning-Based Fast MRI Reconstruction: Improving Generalization for Clinical Translation
1 Introduction
2 Methods
2.1 Background
2.2 Physically-Primed DNN for MRI Reconstruction
3 Experiments
3.1 Dataset
3.2 Experimental Methodology
3.3 Results
4 Conclusions
References
Uncertainty Based Border-Aware Segmentation Network for Deep Caries
1 Introduction
2 Related Work
2.1 Dental Caries Image Segmentation
2.2 Uncertainty Quantification
3 Method
3.1 Border-Aware Network Using SDF
3.2 Uncertainty Based Caries Segmentation
4 Experiments and Discussion
4.1 Dataset and Settings
4.2 Verification of SDF Effectiveness
4.3 Verification of Model Robustness
5 Conclusion
References
An Efficient and Accurate Neural Network Tool for Finding Correlation Between Gene Expression and Histological Images
1 Introduction
2 Methodology
2.1 Data
2.2 Label Generation
2.3 CNN Training and Testing
2.4 Significance Testing and Gene Set Analysis
3 Results
3.1 Resnet Comparison
3.2 Significant Genes and Pathways
3.3 Correlations Between Model Performance and Data Properties
3.4 Comparison of Findings with Other Methodologies
4 Conclusions
References
FAIMI
De-identification and Obfuscation of Gender Attributes from Retinal Scans
1 Introduction
1.1 Differential Privacy for Image Obfuscation
1.2 Deep Learning for Diabetic Retinopathy and Sex Classification
2 Materials and Methods
2.1 Dataset
2.2 Pre-processing
2.3 De-identification Framework
2.4 Evaluation Framework
3 Results
3.1 Full Image Snow Results
3.2 VS-Snow Results
4 Discussion
4.1 Privacy-Utility Tradeoff
4.2 Importance of Vasculature
4.3 Limitations and Future Work
References
Unveiling Fairness Biases in Deep Learning-Based Brain MRI Reconstruction
1 Introduction
2 Background
2.1 Fairness Definitions
2.2 Source of Bias
3 Methods
4 Experimental Analysis
4.1 Dataset and Pre-processing
4.2 Implementation Details
4.3 Results
5 Discussion
6 Conclusion
References
Brain Matters: Exploring Bias in AI for Neuroimaging Research
1 Introduction
2 Current Problems
2.1 Structural Problems
2.2 Specific Biases
3 Mitigation Strategies
3.1 Collect More Representative Data
3.2 Share and Collaborate
3.3 Reduce Reliance on Inaccessible Data Collection Methods
3.4 Develop Both Generic and Specific Models and Employ Transfer Learning
3.5 Consider the Use of Data Augmentation
3.6 Raise Awareness of Bias and Engage in PPI
4 Limitations
5 Conclusion
References
Bias in Unsupervised Anomaly Detection in Brain MRI
1 Introduction
2 Materials and Methods
3 Experiments and Results
3.1 Baseline Performance
3.2 Impact of Bias
3.3 Sources of Bias
4 Conclusion
References
Towards Unraveling Calibration Biases in Medical Image Analysis
1 Introduction
2 Numerical Experiments on Real Data
2.1 Data
2.2 Model Training
2.3 Platt Scaling
2.4 Performance Evaluation
2.5 Results
3 Synthetic Experiments
3.1 Data
3.2 Performance Evaluation
3.3 Results
4 Discussion
References
Are Sex-Based Physiological Differences the Cause of Gender Bias for Chest X-Ray Diagnosis?
1 Introduction
2 Related Work
3 Methods
3.1 Datasets
3.2 Sampling Strategy
3.3 Experimental Settings
4 Results
4.1 Model Performance Across Diseases, Gender Ratios, and Datasets
4.2 Comparison of Different Sampling Strategies
4.3 Breast Cropping Does Not Mitigate Gender Biases
4.4 Dataset Bias v.s. Model Bias
5 Discussion and Conclusions
References
Bayesian Uncertainty-Weighted Loss for Improved Generalisability on Polyp Segmentation Task
1 Introduction
2 Related Work
3 Method
4 Experiments and Results
4.1 Dataset and Experimental Setup
4.2 Results
5 Conclusion
References
Mitigating Bias in MRI-Based Alzheimer's Disease Classifiers Through Pruning of Deep Neural Networks
1 Introduction
2 Materials and Methods
2.1 Data and Preprocess
2.2 Debiasing by Pruning
3 Experiment
3.1 Implementation and Evaluation
3.2 Comparison
4 Result
5 Discussion and Conclusion
References
Auditing Unfair Biases in CNN-Based Diagnosis of Alzheimer's Disease
1 Introduction
2 Materials and Methods
2.1 Data Description and Preprocessing
2.2 Models
2.3 Bias Evaluation Metrics
3 Results and Discussion
3.1 Auditing Fairness with Respect to Model Performance
3.2 Auditing Fairness with Respect to Model Calibration
4 Conclusions
References
Distributionally Robust Optimization and Invariant Representation Learning for Addressing Subgroup Underrepresentation: Mechanisms and Limitations
1 Introduction
2 Assessing Debiasing Mechanisms
2.1 Methodology
2.2 Experiments and Results
3 Improving the Debiasing of Spurious Correlations
References
Analysing Race and Sex Bias in Brain Age Prediction
1 Introduction
2 Materials and Methods
2.1 Bias Analysis
3 Results
4 Discussion and Conclusion
A Appendix
References
Studying the Effects of Sex-Related Differences on Brain Age Prediction Using Brain MR Imaging
1 Introduction
2 Materials and Methods
2.1 Brain MR Datasets
2.2 Pre-processing
2.3 Brain Age Prediction Task
2.4 Grad-CAM Interpretability
2.5 Experimental Setting
3 Results
4 Discussion
5 Conclusion
References
An Investigation into the Impact of Deep Learning Model Choice on Sex and Race Bias in Cardiac MR Segmentation
1 Introduction
2 Materials
3 Methods
3.1 Dataset Sampling
3.2 Model Architecture and Implementation
3.3 Model Evaluation
4 Results
5 Discussion
References
An Investigation into Race Bias in Random Forest Models Based on Breast DCE-MRI Derived Radiomics Features
1 Introduction
2 Materials
3 Methods
4 Experiments and Results
4.1 Race Classification
4.2 Bias Analysis
4.3 Covariate Analysis
5 Discussion and Conclusions
References
How You Split Matters: Data Leakage and Subject Characteristics Studies in Longitudinal Brain MRI Analysis
1 Introduction
2 Methods
2.1 Data Collection and Processing
2.2 Training Setup
2.3 Evaluation Scheme
3 Result
4 Discussion and Conclusion
References
Revisiting Skin Tone Fairness in Dermatological Lesion Classification
1 Introduction
2 Methods and Materials
2.1 Dataset
2.2 Evaluation of Skin Lesion Classification
2.3 Skin Tone Estimation
3 Experiments and Results
3.1 Comparison of ITA Estimation Methods
3.2 Fairness Analysis
3.3 Simulated Data Shifts
4 Conclusions
References
A Study of Age and Sex Bias in Multiple Instance Learning Based Classification of Acute Myeloid Leukemia Subtypes
1 Introduction
2 Materials and Methods
2.1 Data
2.2 Multiple Instance Learning
3 Experiments
3.1 Sex Bias
3.2 Age Bias
4 Results
4.1 Sex Bias
4.2 Age Bias
5 Discussion
References
Unsupervised Bias Discovery in Medical Image Segmentation
1 Introduction
2 Related Work
3 Unsupervised Bias Discovery via Reverse Classification Accuracy
4 Experiments and Discussion
4.1 Synthetic Experiment: Validating RCA for UBD
4.2 Real Experiment: Auditing Chest X-Ray Segmentation Models for Sex Bias
5 Conclusion
References
Debiasing Counterfactuals in the Presence of Spurious Correlations
1 Introduction
2 Methodology
2.1 Classifier Explainability and Debiasing via Counterfactual Image Generation
2.2 Metrics for Evaluating Counterfactuals: Accounting for Spurious Correlations
3 Experiments and Results
3.1 Dataset and Implementation Details
3.2 Results
4 Conclusion
References
EPIMI
On the Relationship Between Open Science in Artificial Intelligence for Medical Imaging and Global Health Equity
1 Introduction
2 Open-Source Data in AI for Medical Imaging
3 Open Science and Transfer Learning
4 Open Science and Distributed Learning
5 Conclusions
References
Gradient-Based Enhancement Attacks in Biomedical Machine Learning
1 Introduction
2 Methods
3 Experiments
4 Discussion
References
Correction to: De-identification and Obfuscation of Gender Attributes from Retinal Scans
Correction to: Chapter 9 in: S. Wesarg et al. (Eds.): Clinical Image-Based Procedures, Fairness of AI in Medical Imaging, and Ethical and Philosophical Issues in Medical Imaging, LNCS 14242, https://doi.org/10.1007/978-3-031-45249-9_9
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken