Twellmann, Thorsten: Data-driven analysis of dynamic contrast-enhanced magnetic resonance imaging data in breast cancer diagnosis. 2005
Inhalt
- Introduction
- Magnetic Resonance Imaging
- Nuclear Magnetism
- Magnetic Resonance Imaging
- Spatial Decomposition of MR Signals
- T1/T2-Weighted Imaging Sequences
- Multispectral Magnetic Resonance Imaging
- Contrast Agents
- Multislice and 3D Imaging
- Summary
- Dynamic Contrast-Enhanced MR Imaging in Breast Cancer Diagnosis
- Anatomy and Disorders of the Breast
- Dynamic Contrast-Enhanced MR Image Sequences
- Interpretation of DCE-MRI Data
- Computer Aided Diagnosis Systems
- Model-Based Image Analysis
- Data-Driven Image Analysis
- Applications of Supervised Artificial Neural Networks
- Applications of Unsupervised Artificial Neural Networks
- Summary
- Supervised Learning - Concepts, Algorithms and Evaluation
- Concepts of Supervised Learning
- Support Vector Machine
- Maximum Margin Hyperplanes
- Kernel Functions
- Hyperparameter Selection
- Multi-Class Extensions
- Output Calibration
- Linear Discriminant Analysis
- Local Sigmoid Map
- Assessment of Classification Performance
- Lesion Detection
- Tissue Characterisation with Artificial Neural Networks
- Motivation
- Data-Driven Pixel-Mapping Based on Supervised Learning
- Setup for a Data-Driven Pixel-Mapping
- Preprocessing of Image Data
- Preparing Training Data
- Adaptation of Multiclass Support Vector Machines
- Adaptation of Local Sigmoid Maps
- Evaluation
- Results
- Discussion
- Adaptive Colour Scales for Comparison of Pixel-Mapping Techniques
- Visualising Pixel-Mapping Functions
- Adaptive Colour Scales
- Low-Dimensional Forms of Signal Spaces Based on Self-Organising Maps
- Self-Organising Maps
- Visualising Adaptive Colour Scales
- Case Study: Comparison of Pixel-Mapping Techniques for DCE-MRI
- Discussion
- Image Fusion for DCE-MRI Data
- Principal Component Analysis and Kernel Principal Component Analysis
- Fusion of DCE-MR Image Sequences
- Preprocessing
- Setup I - Case-Specific Representation Spaces
- Setup II - Domain-Specific Representation Spaces
- Displaying Fused Images
- Results
- Setup I - Case-Specific Representation Spaces
- Setup II - Domain-Specific Representation Spaces Based on PCA
- Setup II - Domain-Specific Representation Spaces Based on KPCA
- ROC Analysis
- Discussion
- Conclusion
- Bibliography