Contents -- Preface -- 1. Biomedical Databases and Data Mining -- 1.1 Databases and Knowledge Discovery in Biomedicine -- 1.2 Outline of this Book -- 2. DYNASTAT: A Methodology for Modeling of Multiagent Systems -- 2.1 Introduction -- 2.2 Literature Review -- 2.2.1 Multi-agent Systems in the Medical Domain -- 2.2.2 Methodologies for Modeling Multi-agent Systems -- 2.2.3 UML Modeling of Multi-agent Systems -- 2.3 DYNASTAT Methodology -- 2.3.1 Conceptual Overview of Multi-agent Systems -- 2.3.2 Modeling Phases -- 2.4 Use of UML 2.2 in the Framework of DYNASTAT Methodology -- 2.4.1 Overview of UML 2.2 -- 2.4.2 Usage of UML for Modeling Multi-agent Systems -- 2.4.3 Selected UML Diagrams -- 2.5 UML to Model Medical Multi-agent Systems -- 2.5.1 Medical Researcher (UML Examples) -- 2.5.1.1 Modeling Information Retrieval -- 2.5.2 Modeling Data Collection and Mining Experiments -- 2.5.3 General Practitioner (UML Examples) -- 2.6 Possible Applications -- 2.7 Conclusion -- 3. SciPort: An Extensible Data Management Platform for Biomedical Research -- 3.1 Introduction -- 3.1.1 Overview -- 3.1.2 Complex Requirements -- 3.1.3 Semantic Consistency -- 3.1.4 Data Sharing -- 3.1.5 Our Contributions -- 3.2 Related Work -- 3.3 Unified Scientific Data Modeling -- 3.3.1 The Scientific Document Model -- 3.3.2 XML Based Implementation of the Data Model -- 3.3.2.1 Metadata -- 3.3.2.2 Content -- 3.3.2.3 The Bene ts of XML -- 3.3.3 XML-based Schema Definition -- 3.3.4 Hierarchical Organization of Data -- 3.4 Document Authoring and Searching -- 3.4.1 Document Authoring -- 3.4.1.1 File Cabinet -- 3.4.1.2 Hierarchy and Schema Authoring -- 3.4.2 DICOM Data Anonymization -- 3.4.3 Searching Documents -- 3.4.4 Semantic Enabled Biomedical Data Management -- 3.4.4.1 Semantic Enabled Data Authoring -- 3.4.4.2 Semantic Enabled Search -- 3.5 Sharing Distributed Biomedical Data.

3.5.1 Sharing Data through a Central Server -- 3.5.2 Data Synchronization -- 3.6 The Architecture of SciPort -- 3.6.1 Open System Architecture -- 3.6.2 Rich Internet Based Application -- 3.6.3 The Architecture Components -- 3.6.4 SciPort Workow -- 3.6.5 High Adaptability through Customization -- 3.7 Conclusion -- 4. An Integrative Framework for Anonymizing Clinical and Genomic Data -- 4.1 Introduction -- 4.1.1 Challenges -- 4.1.2 Contributions -- 4.1.3 Organization -- 4.2 Related Work -- 4.2.1 Data Integration -- 4.2.2 Privacy-preserving Data Sharing -- 4.2.2.1 Anonymization of Di erent Data Types -- 4.2.2.2 Anonymization of Di erent Data Sources -- 4.3 The DIANOVA Framework -- 4.3.1 Setting -- 4.3.1.1 Parties -- 4.3.1.2 Datasets -- 4.3.1.3 Privacy and Utility Requirements -- 4.3.2 Data Flow in DIANOVA -- 4.4 Algorithms for Realizing DIANOVA -- 4.4.1 Algorithms for the Data Integrator (DI) -- 4.4.2 Algorithms for the ANOnymizer (ANO) -- 4.4.3 Algorithms for the View Auditor (VA) -- 4.5 Extensions of DIANOVA -- 4.6 Conclusion -- Acknowledgements -- 5. Data Integration Challenges: A Systems Biology Perspective -- 5.1 Introduction -- 5.2 Modeling Biological Systems -- 5.3 Biological and Mathematical Data Sources -- 5.4 Various Data Exchange Formats in Systems Biology -- 5.5 Building an Integrative Framework to Combine Modeling and Biological Data Sources -- 5.5.1 System Architecture -- 5.5.2 Data Employed in the PDVE -- 5.5.3 Data Employed in Simulations -- 5.5.4 Integration of Protein Interaction Data Sources -- 5.5.5 Integration of Mathematical Modeling Data Sources -- 5.5.6 Database Schema Structure -- 5.5.7 Pathway Model Implementation -- 5.5.8 Exporting the Pathway Model -- 5.6 Analysis of the Developed Integrative Environment -- 5.7 Conclusion -- 6. Ontology-based Data Integration: A Case Study in Clinical Trials -- 6.1 Introduction.

6.2 System Architecture and Overview -- 6.3 The CTDM Ontology -- 6.4 Ontology-based Data Integration of Study Relevant Information -- 6.4.1 Translation between Ontologies and Relational Schemas -- 6.4.2 Mapping Creation and Query Generation -- 6.4.2.1 Mapping of Datatype Properties -- 6.4.2.2 Mapping of Object Properties Representing 1:1-Relationships -- 6.4.2.3 Mapping of Object Properties Representing 1:N-relationships -- 6.4.2.4 Mapping of Object Properties Representing n:m-relationships -- 6.4.2.5 Inheritance Relationships -- 6.5 Assembly of ETL Processes Based on Ontology Mappings -- 6.6 Case Study and Evaluation -- 6.6.1 The Heart Failure Management (HFM) Pilot Trial -- 6.6.2 Evaluation of the CTDMO -- 6.6.3 Evaluation of the Mapping De nition and Integration Module Creation -- 6.6.3.1 Evaluation of Usability -- 6.6.3.2 Evaluation of Feasibility -- 6.7 Related Work -- 6.8 Conclusion -- 7. A Data Warehouse for Ca. Glomeribacter Gigasporarum Bacterium -- 7.1 Introduction -- 7.2 State-of-the-art of Metagenomics for Genomic Comparison -- 7.2.1 GMOD and Chado Database -- 7.3 BIOBITS System Architecture -- 7.3.1 Star Schema in BIOBITS Data Mart -- 7.3.2 System Architecture -- 7.3.2.1 Services on Chado and the Star Schema -- 7.4 Software Modules to Support Researchers' Activities -- 7.4.1 Case-based Reasoning -- 7.4.1.1 Multiple Abstractions on the Genome -- 7.4.1.2 Multi-dimensional Index Structures -- 7.4.2 Clustering Modules -- 7.4.2.1 Co-clustering -- 7.4.2.2 Proximity Measures -- 7.5 Conclusion -- 8. Quality of Medical Data: A Case Study -- 8.1 Introduction -- 8.1.1 Quality and Medical Data -- 8.1.2 Features of Medical Data -- 8.1.3 A Quality Methodology for Medical Data -- 8.1.4 Dimensions of Analysis -- 8.1.4.1 Data-related Dimensions -- 8.1.4.2 Meta-data-related Dimensions -- 8.1.4.3 System-related Dimensions -- 8.2 Case Study.

8.2.1 Description and Extraction of the Data -- 8.2.2 Quality Assessment -- 8.2.2.1 Data -- 8.2.2.2 Meta-data -- 8.2.2.3 System -- 8.3 Generation of a Summary Table -- 8.4 Conclusion -- 9. Efficient EMD-based Similarity Search in Medical Image Databases -- 9.1 Introduction -- 9.1.1 Related Work -- 9.1.2 Formal Definition of the Earth Mover's Distance -- 9.1.3 Multi-Step Query Processing and the EMD -- 9.2 Dimensionality Reduction for the EMD -- 9.2.1 Dimensionality Reduction -- 9.2.2 Optimal Dimensionality Reduction -- 9.2.2.1 Optimal Cost Reduction -- 9.2.2.2 Optimal Flow Reduction -- 9.2.3 Flow-based Reduction -- 9.3 Query Processing Algorithm -- 9.4 Evaluation on Medical Data Sets -- 9.5 Conclusion -- 10. Fast Multimedia Querying for Medical Applications -- 10.1 Introduction -- 10.2 Related Work -- 10.3 Subspace Tree -- 10.3.1 Orthogonal Projection and the Mean Value -- 10.3.2 Image Pyramid - A Subspace Tree -- 10.4 Experiments -- 10.5 Conclusion -- Acknowledgments -- 11. Ensemble Feature Selection in Biomedical Applications -- 11.1 Introduction -- 11.1.1 Statistical Hypothesis Testing -- 11.1.2 Information Gain -- 11.1.3 ReliefF -- 11.1.4 Biomarker Identifier -- 11.2 Evaluation of Feature Selection Approaches -- 11.2.1 Popular Classifiers for Assessing Discriminatory Ability of Selected Features -- 11.2.2 Classifier Validation -- 11.3 Ensemble Feature Selection -- 11.3.1 Stacked Feature Ranking -- 11.4 Biomedical Example -- 11.4.1 Data Collection -- 11.4.2 Sample Preparation and Mass Spectrometry Analysis -- 11.5 Computational Approach -- 11.5.1 Experimental Setup -- 11.6 Results -- 11.6.1 Comparison of the Predictive Power of the Applied Feature Ranking Methods -- 11.6.2 Identified Breath Gas Marker Candidates -- 11.7 Discussion -- 11.8 Conclusion -- Acknowledgment -- 12. Analysis of Breast Cancer Genomic Data by Fuzzy Association Rule Mining.

12.1 Introduction -- 12.1.1 Breast Cancer -- 12.1.1.1 Tumor stage -- 12.1.1.2 HER2/neu (Human Epidermal Growth Factor Receptor 2) -- 12.1.1.3 HER2 Testing Methodologies -- 12.1.1.4 Additional Biomarkers -- 12.2 Microarrays -- 12.3 Association Rule Mining -- 12.3.1 Association Rules: Formal Definition -- 12.4 Fuzzy Association Rules -- 12.5 Dataset -- 12.5.1 HER2 Testing Methodologies -- 12.5.2 Immunohistochemical Data -- 12.5.3 Microarray Data -- 12.6 Extracting the Fuzzy Association Rules -- 12.6.1 Fuzzy Top-Down Frequent-Pattern Growth Algorithm -- 12.6.1.1 Obtaining the List of Frequent Items -- 12.6.1.2 Building the Fuzzy Frequent-pattern Tree -- 12.6.1.3 Frequent Itemset Generation -- 12.6.2 Obtaining and Processing the Fuzzy Association Rules -- 12.6.2.1 Generating the Fuzzy Association Rules from the Frequent Itemsets -- 12.6.2.2 Deficiencies of the Con dence/Support Framework -- 12.7 Results -- 12.7.1 Exploratory Analysis of 2751 Patients -- 12.7.2 Whole-genome Expression Data and Prognostic Factors -- 12.8 Conclusion -- Acknowledgements -- 13. Graph Mining on Brain Co-activation Networks -- 13.1 Introduction -- 13.2 Related Work -- 13.2.1 Graph Dataset Mining -- 13.2.2 Large Graph Mining -- 13.3 Method -- 13.3.1 Basics of Graph Theory -- 13.3.2 Construction of Brain Co-activation Networks Out of fMRI Timeseries -- 13.3.3 Performing Frequent Subgraph Mining on Brain Co-activation Networks -- 13.3.4 Evaluation of Detected Motifs -- 13.4 Experiments -- 13.5 Conclusion -- 14. Automatic Identification of Surgery Indicators -- 14.1 Introduction -- 14.1.1 Aims and Objectives -- 14.1.2 Outline -- 14.2 Background -- 14.2.1 Referral Contents -- 14.2.2 Related Work -- 14.3 Approach -- 14.3.1 Supervised Learning -- 14.3.2 Theoretical Modeling -- 14.3.3 Models for Structured Referrals -- 14.3.4 Prediction Models for the Surgical Suite.

14.3.5 Algorithm Selection.