Contents
Foreword
Acknowledgments
1. Introduction to Tissue Phenomics
Ralf Huss, Gerd Binnig, Günter Schmidt, Martin Baatz and Johannes Zimmermann
1.1 Motivation
1.2 Tissue Phenomics and Medicine 4.0
1.3 Phenes in Cancer Immunology
1.4 Future of Tissue Phenomics
1.5 About This Book
2. Image Analysis for Tissue Phenomics
Johannes Zimmermann, Keith E. Steele, Brian Laffin, René Korn, Jan Lesniak, Tobias Wiestler and Martin Baatz
2.1 Introduction
2.2 Experimental Design for Image Analysis Studies
2.3 Input Data and Test Data
2.4 ImageAnalysis
2.5 Quality Control Procedures
2.6 Results and Output
2.7 Discussion and Summary
3. Context-Driven Image Analysis: Cognition Network Language
Gerd Binnig
3.1 Motivation and Reasoning
3.2 History and Philosophy
3.3 Technology
3.3.1 Class Hierarchy
3.3.2 Process Hierarchy and Context Navigation
3.3.2.1 Processes and the domain concept
3.3.2.2 Context navigation
3.3.2.3 Object-based processing
3.3.2.4 Maps
3.3.2.5 Object variables
3.4 Example of Context-Driven Analysis
3.4.1 H&E Analysis Problems
3.4.2 Concrete H&E Image Analysis Example
3.4.2.1 Color channel analysis as a starting point
3.4.2.2 Isolated nuclei: first context objects
3.4.2.3 Some remaining large nuclei by splitting and insideout procedure
3.4.2.4 Nuclei of immune cells
3.4.2.5 Template match for heterogeneous nuclei on 10x and 20x
3.5 Conclusion
4. Machine Learning: A Data-Driven Approach to Image Analysis
Nicolas Brieu, Maximilian Baust, Nathalie Harder, Katharina Nekolla, Armin Meier and GünterSchmidt
4.1 Introduction: From Knowledge-Driven to Data-Driven Systems
4.2 Basics of Machine Learning
4.2.1 Supervised Learning
4.2.2 Classification and Regression Problems
4.2.3 Data Organization
4.3 Random Forests for Feature Learning
4.3.1 Decision Trees
4.3.1.1 Definition
4.3.1.2 Decision function
4.3.1.3 Extremely randomized trees
4.3.1.4 Training objective function
4.3.2 Random Forests Ensemble (Bagging)
4.3.3 Model Parameters
4.3.4 Application Generic Visual Context Features
4.3.4.1 Haar-like features
4.3.4.2 Gaborfeatures
4.3.5 Application to the Analysis of Digital Pathology Images
4.3.5.1 On-the-fly learning of slide-specific random forests models
4.3.5.2 Area-guided distance-based detection of cell centers
4.4 Deep Convolutional Neural Networks
4.4.1 History: From Perceptrons and the XOR Problem to Deep Networks
4.4.2 Building Blocks
4.4.2.1 Convolutional layers
4.4.2.2 Convolutional neural networks
4.4.2.3 Loss functions
4.4.2.4 Activation functions
4.4.2.5 Pooling layers
4.4.2.6 Dropout layers
4.4.3 Application Examples
4.5 Discussion and Conclusion
4.5.1 Model Complexity and Data Availability
4.5.2 Knowledge and Data Teaming
4.5.3 Machine Learning Teaming
4.5.4 Conclusion
5. Image-Based Data Mining
Ralf Schönmeyer, Arno Schäpe and Günter Schmidt
5.1 Introduction
5.2 Generating High-Level Features for Patient Diagnosis
5.2.1 Quantification of Regions of Interest
5.2.2 Heatmaps
5.2.2.1 of heatmaps
5.2.2.2 Visualization of heatmaps
5.2.2.3 Heatmaps with multiplexed data
5.2.2.4 Objects in heatmaps
5.2.3 Algebraic Feature Composition
5.2.4 Aggregation of Multiple Tissue Pieces to Single Patient Descriptors
5.2.5 Integration of Clinical Data and Other Omics Data
5.3 Performance Metrics
5.4 Feature Selection Methods
5.4.1 Unsupervised Methods
5.4.2 Hypothesis-Driven Methods
5.4.3 Univariate, Data-Driven Feature Selection
5.5 Tissue Phenomics Loop
5.6 Tissue Phenomics Software
5.6.1 Image Analysis and Data Integration
5.6.2 General Architecture
5.6.3 Image Mining Software
5.6.4 Data and Workflow Management and Collaboration
5.7 Discussion and Outlook
6. Bioinformatics
Sriram Sridhar, Brandon W. Higgs and Sonja Althammer
6.1 Molecular Technologies: Past to Present
6.2 Genomics and tissue phenomics
6.2.1 Genomics data sources
6.2.2 The art of image mining
6.2.2.1 Cell-to-cell distances
6.2.2.2 Quantifying cell populations in histological regions
6.2.2.3 Tissue phene and survival
6.2.3 Power of integrative approaches
6.3 Analytical approaches for image and genomics data
6.3.1 Data handling requires similar methods
6.3.2 Enhancing confidence in a discovery
6.4 Examples of genomics or IHC biomarkers in clinical practice
6.4.1 Biomarker background
6.4.2 Genomics and IHC to guide prognosis or diagnosis
6.4.3 Patient stratification: Genomics and IHC to identify patient subsets for treatment
7. Applications of tissue phenomics
Johannes Zimmermann, Nathalie Harder and Brian Laffin
7.1 Introduction
7.2 Hypothesis-Driven approaches
7.2.1 TME as a battlefield: CD8 and PD-L1 densities facilitate patient stratification for durvalumab therapy in non-small cell lung cancer
7.2.2 Gland morphology and TAM distribution patterns outperform gleason score in prostate cancer
7.2.3 Novel spatial features improve staging in colorectal cancer
7.2.4 Immunoscore is a novel predictor of patient survival in colorectal cancer
7.2.5 Analysis of spatial interaction patterns improves conventional cell density scores in breast cancer
7.2.6 Immune cell infiltration is prognostic in breast cancer
7.2.7 The immune landscape structure directly corresponds with clinical outcome in clear-cell renal cell carcinoma
7.3 Hypothesis-Free prediction
7.4 Summary
8. Tissue Phenomics For Diagnostic Pathology
Maria Athelogou and Ralf Huss
8.1 Introduction
8.2 Digital Pathology
8.3 Tissue Phenomics Applications for Computer-Aided Diagnosis in Pathology
8.3.1 Technical Prerequisites for Tissue Phenomics
8.4 Tissue Phenomics Applications for Decision Support Systems in Pathology
8.5 Summary of Pathology 4.0
9. Digital Pathology: Path into the Future
Peter D. Caie and David J. Harrison
9.1 Introduction
9.2 Brief History of Digital Pathology
9.3 Digital Pathology in Current Practice
9.3.1 Research
9.3.2 Education
9.3.3 Clinical
9.4 Future of Digital Pathology
9.4.1 Mobile Scanning
9.4.2 Feature-Based Image Analysis
9.4.3 Machine Learning on Digital Images
9.4.4 Big Data and Personalized Pathology
10. Tissue Phenomics in Clinical Development and Clinical Decision Support
Florian Leiß and Thomas Heydler
10.1 Cancer and oncology drug development
10.2 Immunotherapy
10.3 Digitalization of healthcare
10.4 Patient profiling
10.5 Therapy matching
10.6 Benefits
10.7 Conclusion
Glossary
Index
Foreword
Evolution of Tissue Phenomics and Why It Is Critical to the War on Cancer
A view from a tumor immunologist and cancer immunotherapist
Those of us in biomedical research are witnessing an almost daily evolution of our science. Nowhere is this more obvious, or possessing greater impact, than in the field of cancer immunology and immunotherapy. Cancer, one of the great scourges on humanity, is having the veil of its secrets lifted. Digital imaging and objective assessment tools contribute substantial and solid evidence to document that immune cells are prognostic biomarkers of improved outcomes for patients with cancer. While anecdotal reports of associations between immune cell infiltrates and improved outcomes have been presented by pathologists for more than 100 years, the co-evolution of multiple science subspecialties has resulted in opportunities to better understand the disease and why it develops. Armed with this knowledge and evidence that checkpoint blockade therapies are capable of unleashing the immune system, increasing survival and possibly curing some patients with cancer, will lead to additional investment in this area of research, which will accelerate the pace at which we develop improved treatments for cancer. It is clear that digital imaging and assessment of complex relationships of cells within cancer, the very essence of tissue phenomics will play a central role in the development of the next generation of cancer immunotherapies. Ultimately, assessment of cancer tissue phenomics will be used to tailor immunotherapies to treat and eventually cure patients with cancer.
This is a very different time from when I began as an immunologist. Monoclonal antibodies, reagents capable of objectively assessing thousands of molecules, were not yet invented. To identify a subset of white blood cells, termed T cells, we used the bind...