Interpretability of Machine Intelligence in Medical Image Computing : 5th International Workshop, iMIMIC 2022, Held in Conjunction with MICCAI 2022, Singapore, Singapore, September 22, 2022, Proceedings / edited by Mauricio Reyes, Pedro Henriques Abreu, Jaime Cardoso.

Format:

Book

Contributor:

Henriques Abreu, Pedro, editor.

Cardoso, Jaime, editor.

Reyes, Mauricio, editor.

Series:

Lecture Notes in Computer Science, 1611-3349 ; 13611

Language:

English

Subjects (All):

Computer vision.

Machine learning.

Education--Data processing.

Education.

Social sciences--Data processing.

Social sciences.

Bioinformatics.

Computer Vision.

Machine Learning.

Computers and Education.

Computer Application in Social and Behavioral Sciences.

Computational and Systems Biology.

Local Subjects:

Computer Vision.

Machine Learning.

Computers and Education.

Computer Application in Social and Behavioral Sciences.

Computational and Systems Biology.

Physical Description:

1 online resource (134 pages)

Edition:

1st ed. 2022.

Place of Publication:

Cham : Springer Nature Switzerland : Imprint: Springer, 2022.

Summary:

This book constitutes the refereed joint proceedings of the 5th International Workshop on Interpretability of Machine Intelligence in Medical Image Computing, iMIMIC 2022, held in September 2022, in conjunction with the 25th International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2022. The 10 full papers presented at iMIMIC 2022 were carefully reviewed and selected from 24 submissions each. The iMIMIC papers focus on introducing the challenges and opportunities related to the topic of interpretability of machine learning systems in the context of medical imaging and computer assisted intervention. .

Contents:

Intro

Preface

Organization

Contents

Interpretable Lung Cancer Diagnosis with Nodule Attribute Guidance and Online Model Debugging

1 Introduction

2 Materials

3 Methodology

3.1 Collaborative Model Architecture with Attribute-Guidance

3.2 Debugging Model with Semantic Interpretation

3.3 Explanation by Attribute-Based Nodule Retrieval

4 Experiments and Results

4.1 Implementation

4.2 Quantitative Evaluation

4.3 Trustworthiness Check and Interpretable Diagnosis

5 Conclusions

References

Do Pre-processing and Augmentation Help Explainability? A Multi-seed Analysis for Brain Age Estimation

2 Related Work

3 Methods

4 Results

4.1 Performance

4.2 Voxel Agreement

4.3 Atlas-Based Analyses

4.4 Region Validation

5 Conclusion

Towards Self-explainable Transformers for Cell Classification in Flow Cytometry Data

3.1 Architecture

3.2 Preprocessing

3.3 Loss Function

3.4 Data Augmentation

4 Experiments

4.1 Data

4.2 Results

Reducing Annotation Need in Self-explanatory Models for Lung Nodule Diagnosis

2 Method

3 Experimental Results

3.1 Prediction Performance of Nodule Attributes and Malignancy

3.2 Analysis of Extracted Features in Learned Space

3.3 Ablation Study

4 Conclusion

Attention-Based Interpretable Regression of Gene Expression in Histology

2 Methods

2.1 Datasets

2.2 Multiple Instance Regression of Gene Expression

2.3 Attention-Based Model Interpretability

2.4 Evaluation of Performance and Interpretability

3 Experiments and Results

3.1 Network Training

3.2 Quantitative Model Evaluation

3.3 Attention-Based Identification of Hotspots and Patterns.

3.4 Quantitative Evaluation of the Attention

4 Discussion

A Description of Selected Genes

B Detailed Model Evaluation

C Additional Visualizations

D Single-Cell Co-expression

Beyond Voxel Prediction Uncertainty: Identifying Brain Lesions You Can Trust

2 Our Framework: Graph Modelization for Lesion Uncertainty Quantification

2.1 Monte Carlo Dropout Model and Voxel-Wise Uncertainty

2.2 Graph Dataset Generation

2.3 GCNN Architecture and Training

3 Material and Method

3.1 Data

3.2 Comparison with Known Approaches

3.3 Evaluation Setting

3.4 Implementation Details

4 Results and Discussion

Interpretable Vertebral Fracture Diagnosis

1.1 Related Work

2 Methodology

2.1 Vertebral Fracture Detection

2.2 Semantic Concept Extraction (Correlation)

2.3 Visualization of Highly Correlating Concepts at Inference

3 Experimental Setup

4.1 Clinical Meaningfulness of Extracted Semantic Concepts

4.2 Single-Inference Concept Visualization

Multi-modal Volumetric Concept Activation to Explain Detection and Classification of Metastatic Prostate Cancer on PSMA-PET/CT

2 Data

3 Method

3.1 Preprocessing

3.2 Detection

3.3 Classification

3.4 Explainable AI

4.1 Detection

4.2 Classification

4.3 Explainable AI

5 Discussion

6 Conclusion

KAM - A Kernel Attention Module for Emotion Classification with EEG Data

3 Kernel Attention Module

Explainable Artificial Intelligence for Breast Tumour Classification: Helpful or Harmful

2.1 XAI in Medicine.

3 Model Setup

3.1 Data Pre-Processing

3.2 Model Architecture

4 Explanations

4.1 LIME

4.2 RISE

4.3 SHAP

5 Evaluating Explanations

5.1 One-Way ANOVA

5.2 Kendall's Tau

5.3 Radiologist Evaluation

5.4 Threats to Validity

6 Observations and Discussion

6.1 Discussion

A Appendix

A.1 Model Training Results

A.2 Choosing L Parameter for LIME

A.3 One-Way ANOVA Results

A.4 Pixel Agreement Statistics

A.5 Ranked Biased Overlap (RBO) Results

A.6 Kendall's Tau Results

A.7 Radiologist Opinions

A.8 Explanation Examples

Author Index.

Notes:

Includes index.

Other Format:

Print version: Reyes, Mauricio Interpretability of Machine Intelligence in Medical Image Computing

ISBN:

9783031179761

3031179765

OCLC:

1492949481