Artificial Intelligence and Data Mining in Healthcare

Preface
	Need for a Book on the Proposed Topics
	Organization of the Book
	Audience
Contents
Contributors
Acronyms
1 Artificial Intelligence for Healthcare Logistics: An Overview and Research Agenda
	1.1 Introduction
	1.2 Machine Learning and Artificial Intelligence
		1.2.1 Machine Learning
		1.2.2 Artificial Intelligence
		1.2.3 Working Definition
	1.3 Framework for Healthcare Logistics Literature
		1.3.1 Planning Levels
		1.3.2 Care Levels
		1.3.3 User Types
		1.3.4 Framework
	1.4 Literature Review
		1.4.1 AI for Optimisation Input
		1.4.2 AI for Healthcare Logistics Optimisation
		1.4.3 AI for ED Logistics
		1.4.4 Synthesis and Research Agenda
	1.5 Conclusion
	References
2 AI/OR Synergies of Process Mining with Optimal Planning of Patient Pathways for Effective Hospital-Wide Decision Support
	2.1 Motivation and Research Outline
		2.1.1 AI/OR Synergies meet Hospital Decision Task Complexities
		2.1.2 Pathway Centered Decision Support Toward AI/OR Synergy
		2.1.3 Research on AI/OR Synergy and Chapter Outline
	2.2 First Type of AI/OR Synergy: Process Mining of Pathways for Accurate Prescriptive Planning of Ward-and-Bed Allocation
		2.2.1 Synergy between Predictive and Prescriptive Analytics: Cases of Simple vs. Complex Structures
		2.2.2 First Type of AI/OR Synergy and Its Benefits for Effective Hospital Decision Support: Case Study of a University Hospital
	2.3 Detecting AI/OR Synergies Within Hospital Decision Support: Interdependencies, Dimensions of Complexity, Two-Dimensional Scheme, and Types of AI/OR Synergy
		2.3.1 Types of Interdependencies: First Group
		2.3.2 Dimensions of Complexity and Overview About OR and AI Tasks and Synergies
		2.3.3 A New Two-Dimensional Scheme for Simulation-/Optimization-Based Decision Support in Hospitals Applied to Overall Bed Management in Interdependent Wards
		2.3.4 AI Tasks and AI/AI Synergy: Stepwise Aggregation from Process Mining to More Accurate Hospital Data Mining
		2.3.5 OR Tasks and OR/OR Synergies
		2.3.6 First Type of AI/OR Synergy and Detecting a Second Type
	2.4 Second Type of AI/OR Synergy: Mining of Process Discrepancies and Its Interplay with Prescriptive Planning Toward Effective Hospital-Wide Decision Support
		2.4.1 Types of Interdependencies: Second Group and Model–Reality Gap
		2.4.2 Mining Process Discrepancies by Type of Interdependency
		2.4.3 Interplay Between Mining Process Discrepancies with Prescriptive Planning and Operationalization of the Second Type of AI/OR Synergy by a Discrepancy-Driven Approach
	2.5 Conclusion
	References
3 Real-Time Capacity Management and Patient Flow Optimization in Hospitals Using AI Methods
	3.1 Introduction
	3.2 Capacity Management in Hospitals
		3.2.1 Traditional Hospital Capacity Management
		3.2.2 Queuing and Synchronization in Hospitals
	3.3 AI Methods for Hospital Capacity Management
		3.3.1 Multi-Agent Systems
		3.3.2 Artificial Neural Networks (ANN)
		3.3.3 Machine Learning (ML)
	3.4 Example of AI in Capacity Management and Patient Flow Optimization
		3.4.1 Methods
	3.5 Conclusion
	3.6 Future of AI in Patient Flow Optimization and Capacity Management
	References
4 How the Health-Care Expenditure Influences the Life Expectancy: Case Study on Russian Regions
	4.1 Introduction
	4.2 Life Expectancy as a Symbolic Regression Problem
	4.3 Variable Neighborhood Programming for Solving Symbolic Regression Problem
	4.4 Case Study on Life Expectancy at Russian Districts
		4.4.1 One-Attribute Analysis
		4.4.2 Results and Discussion on Three-Attribute Data
	4.5 Conclusions
	References
5 Operating Theater Management System: Block-Scheduling
	5.1 General Context
		5.1.1 Introduction
		5.1.2 CHU Operating Theater ``Block-Schedule''
		5.1.3 Search Background
		5.1.4 Problem Definition
	5.2 MILP Problem Formulation
		5.2.1 Definition of Decision Variables
		5.2.2 Objective Function
		5.2.3 Constraints
		5.2.4 Results of the Simulation
	5.3 MAS Planner Approach
		5.3.1 Preface
		5.3.2 Multi-Agent Planner
		5.3.3 Patient's Programming
		5.3.4 Frequency Evaluation
		5.3.5 Surgeon's Preferences
		5.3.6 Virtual Cost
		5.3.7 Block-Scheduling Algorithm
		5.3.8 Optimization Algorithm
		5.3.9 Performance Metrics
	5.4 Experimental Test
		5.4.1 Test Data
		5.4.2 Simulation Results
	5.5 Conclusion
	References
6 An Immune Memory and Negative Selection to Visualize Clinical Pathways from Electronic Health Record Data
	6.1 Introduction
	6.2 What is an EHR?
		6.2.1 Benefits with the EHR
		6.2.2 Better Practice Management with the EHR
	6.3 Current Study
	6.4 Overview of the System
		6.4.1 Representation of the Self-Cell
		6.4.2 Antigen Representation
		6.4.3 Representation of B-Cells
	6.5  Negative Selection Algorithm for System Monitoring
		6.5.1 Step 1: Learning
		6.5.2 Step 2: Monitoring
	6.6 Control of EHRs by Memory Cells
		6.6.1 The Algorithm Developed by Immune Memory (IMA)
	6.7 Implementation and Results
	6.8 Conclusion
	References
7 Optimized Medical Image Compression for Telemedicine Applications
	7.1 Introduction
	7.2 Previous Approaches
	7.3 Preliminaries
		7.3.1 Legendre Moments
		7.3.2 Whale Optimization Algorithm (WOA)
			7.3.2.1 Encircling Prey
			7.3.2.2 Bubble-Net Attacking Method (Exploitation Phase)
			7.3.2.3 Search for Prey (Exploration Phase)
	7.4 The Proposed Compression Method
	7.5 Numerical Experiments
		7.5.1 Test image
		7.5.2 Performance Measures
		7.5.3 Results and Discussion
	7.6 Limitations of the Proposed Algorithm
	7.7 Conclusion
	References
8 Online Variational Learning Using Finite Generalized Inverted Dirichlet Mixture Model with Feature Selection on Medical Data Sets
	8.1 Introduction
	8.2 Clustering Applications in Healthcare
	8.3 Model Specification
		8.3.1 Finite Generalized Inverted Dirichlet Mixture Model with Feature Selection
		8.3.2 Prior Specifications
	8.4 Online Variational Learning for Finite Generalized Inverted Dirichlet Mixture Mode with Feature Selection
		8.4.1 Variational Inference
		8.4.2 Online Variational Inference
	8.5 Experimental Results
		8.5.1 Image Segmentation
		8.5.2 Synthetic Data
		8.5.3 Medical Image Data Sets
			8.5.3.1 Brain Tumor Detection
			8.5.3.2 Skin Melanoma Detection
			8.5.3.3 Malaria Data Set
	8.6 Conclusion
	Appendix
		Update the Variational Hyper-Parameters
	References
9 Entropy-Based Variational Inference for Semi-Bounded Data Clustering in Medical Applications
	9.1 Introduction
	9.2 Finite Inverted Dirichlet Mixture Model
	9.3 Entropy-Based Variational Learning
		9.3.1 Variational Learning
		9.3.2 Model Learning Through Entropy-Based Variational Bayes
		9.3.3 Theoretical Entropy of Inverted Dirichlet Mixtures
		9.3.4 MeanNN Entropy Estimator
	9.4 Experimental Results
		9.4.1 Cardiovascular Diseases (CVDs)
		9.4.2 Diabetes
		9.4.3 Lung Cancer
		9.4.4 Breast Cancer
	9.5 Conclusion
	References