AI-Driven Smart Healthcare

Ajay Pratap is an Assistant Professor in the Department of Computer Science and Engineering at the Indian Institute of Technology (BHU), Varanasi, India. Yashwant Singh Patel is a Postdoctoral Fellow in the Department of Computing Science in Umeå University, Sweden and an Assistant Professor with the Computer Science and Engineering Department (CSED) at Thapar Institute of Engineering & Technology, Punjab, India. Ram Narayan Yadav is an Assistant Professor in the Electrical and Computer Science Engineering department at the Institute of Infrastructure Technology Research and Management in Ahmedabad, India. Ali Ahmadian is a Senior Research Scientist at the Decisions Lab, Mediterranean University of Reggio Calabria, Italy and an Adjunct Professor with the Faculty of Engineering and Natural Sciences, Istanbul Okan University, Istanbul, Turkey. Ashok Kumar Yadav is an Assistant Professor in the Department of Information Technology at the Rajkiya Engineering College in Azamgarh, Uttar Pradesh, India.

• Notes on Contributors xvPreface xviiAcknowledgments xixList of Abbreviations xxiIntroduction xxv1 Internet of Things, Edge, Fog, and Data Analytics in Smart Healthcare: Introduction, Benefits, and Challenges 11.1 Introduction 11.2 Use of Edge and Fog Computing for Healthcare Applications 31.2.1 Role of Edge Computing in Resource Management 71.3 Data Analytics in Healthcare Applications 81.3.1 Components of BDA 91.4 BDA Applications 101.4.1 Challenge in Utilization of Big Data in Healthcare 111.5 Use Case Scenarios 131.5.1 Use of Data Analysis for Forecasting TB Prevalence Rates 131.5.2 Skin Aging Estimation 141.5.3 Other Use Cases in Smart Healthcare 161.6 Current Challenges and Future Directions 171.7 Future Direction: Opening Up Health Data for Research 191.8 Conclusion 20References 212 Hyperscale Computing Paradigm in Healthcare 292.1 Introduction 292.2 The Evolution of Computing in Healthcare 302.2.1 Early Era of Digitalization 302.2.2 Transition to Cloud Computing 312.3 What Is Hyperscale Cloud? 322.4 Components of Hyperscale Computing 322.4.1 Distributed Systems 322.4.2 Cloud Infrastructure 342.4.3 Load Balancers 352.4.4 Storage Systems 362.4.5 Networking and Interconnectivity 362.4.6 Compute Resources 362.4.7 Automation and Orchestration 372.4.8 Security and Compliance 372.5 Challenges of Hyperscale Computing in Healthcare 372.6 Hyperscale Data Centers 392.6.1 Key Components of a Hyperscale Data Center 392.7 Tech Giants Playing a Role in Hyperscaling 402.8 Public Versus Private Hyperscale Clouds in Healthcare 422.9 Depth and Future of Hyperscale Computing in Healthcare 422.9.1 Integration of Connected Healthcare and Hyperscale Cloud 442.10 Case Studies of Hyperscale Computing in Healthcare 45References 473 Containerized Internet of Medical Things and Serverless Computing for Smart Healthcare Systems 493.1 Wireless Technologies Empowering Internet of Medical Things 493.2 IoMT and Its Role in Modern Healthcare 503.3 Importance of Containerization 523.3.1 Benefits of Containerization 523.3.2 Use Cases of Containerization 533.4 Serverless Computing: Concept and Overview 543.4.1 Features of Serverless Computing 553.5 Complementing Containerization for Healthcare Applications 553.6 Real-world Use Cases 573.7 Future Directions in Containerized IoMT and Serverless Healthcare 583.8 Conclusion 58References 594 Kubernetes Enabled Resource Management Architecture for Secure Innovation in Healthcare 614.1 Overview of Kubernetes and Its Role in Healthcare 614.2 Key Features of Kubernetes 624.3 Key Kubernetes Concepts 634.4 Kubernetes Control Plane and Nodes 644.5 Kubernetes Resource Management 654.6 Benefits of Kubernetes for the Healthcare Industry 664.7 Use Cases of Cloud-native and Kubernetes in Healthcare 674.8 Kubernetes as a Solution 684.9 Conclusion 70References 715 Exploring Artificial Intelligence (AI) and Machine Learning (ML) for Performance and Predictive Analysis of Various Diseases Using Health-related Data 735.1 Introduction 735.2 Challenges in Healthcare 755.3 Significance of AI and ML in Healthcare 765.3.1 Early Diagnosis and Disease Detection 765.3.2 Personalized Medicine 775.3.3 Improving Clinical Decision Support 775.3.4 Reducing Administrative Burden 775.3.5 Predicting and Managing Disease Outbreaks 775.3.6 Enhancing Drug Discovery and Development 775.4 Application of AI/ML in Healthcare System 785.4.1 Prognosis 795.4.2 Diagnosis 805.4.3 Treatment 805.4.4 Clinical Workflows 805.5 Major Development Phases of AI/ML-based Healthcare Systems 815.5.1 Use Case Specification 835.5.2 Data Access and Anonymization 835.5.3 Data Annotation 835.5.4 Model Development 835.5.5 Model Testing and Auditing 845.5.6 Multi-site Verification and Validation 845.5.7 Regulatory Approvals 845.5.8 Clinical Integration 845.5.9 User Acceptance 855.5.10 Real-world Surveillance 855.6 Secure, Private, and Robust AI/ML-based Healthcare: Challenges 855.6.1 Vulnerabilities in Data Collection 855.6.2 Vulnerabilities Due to Data Annotation 865.6.3 Vulnerabilities in Model Training 865.6.4 Vulnerabilities in Deployment Phase 875.6.5 Vulnerabilities in Testing Phase 875.7 Use Case: Diabetes 875.7.1 Predictive Analysis in Disease Management 895.7.2 Performance Analysis in Healthcare 895.8 Challenges and Future Directions 935.9 Conclusion 93References 986 Algorithmic Frameworks for Cost Minimization in Criticality Aware Mobile Healthcare System 1016.1 Introduction 1016.2 Related Works 1026.3 System Model 1046.3.1 Data Criticality Index 1066.4 Problem Definition 1086.4.1 Auction Process 1086.4.2 LDPU Perspective 1106.4.3 CSP Perspective 1106.4.4 Mechanism Perspective 1106.4.5 Impact of Selfishness 1116.5 Proposed Auction Mechanism 1116.5.1 Cheating Detection Process at LDPU 1126.5.2 Decision Process at CSP 1136.5.3 Cheating Detection Process at CSP 1136.5.4 Decision Process at LDPU 1146.5.5 An Illustrative Example 1146.6 Analysis of Proposed Mechanism 1176.6.1 Truthful Analysis at LDPU Side 1176.6.2 Truthful Analysis at CSP Side 1186.7 Performance Study 1206.8 Conclusion 124References 1247 Utility-aware Edge Computing System for Remote Health Monitoring 1277.1 Introduction 1277.2 Related Works 1297.3 System Model and Problem Formulation 1317.3.1 Reputation Scheme 1327.4 Proposed Auction Model 1367.4.1 Auction Details 1397.5 Analysis of Proposed Auction Model 1427.6 Performance Evaluation 1437.6.1 Individual Rationality 1447.6.2 Budget Balance 1457.6.3 Utilities of Model Owners and Users 1467.7 Conclusion 148References 1488 Fog Computing-based WBAN and IoT Framework for Prediction of Various Diseases Using Big Data Analytics 1518.1 Introduction 1518.2 The Role of Fog Computing in Healthcare 1538.3 Big Data Analytics in Healthcare 1538.4 Wireless Body Area Networks in Healthcare 1558.5 Framework Design 1568.6 Disease Prediction Use Case for Healthcare 1588.7 Conclusion 163References 1639 Disease Spread Detection and Controlling with Fog-based Model in Wireless Body Area Networks 1679.1 Compartmental Model 1699.1.1 Applications of Compartmental Models 1719.1.2 Contact-based Models 1729.2 Simulation Tools 1749.3 Other Environmental Factors to Control Spread of Disease 1779.4 AI- and ML-based Health Prediction Approaches 1789.5 Conclusion 180References 18010 Optimized Doctor Recommendation System Using Machine Learning Approach 18510.1 Introduction 18510.2 Related Works 18710.3 System Model 19010.3.1 Selection Factor of Doctor 19110.3.2 Expertise Score of Doctor 19210.3.3 Expected Reward of Doctor 19210.3.4 Satisfaction Level of Patient 19410.3.5 Problem Formulation 19410.4 Proposed Approach 19510.4.1 Calculate Expertise Score 19610.4.2 Recommend an Ordered Assortment 19810.4.3 Update Doctor’s Information 19910.5 Performance Study 20110.6 Conclusion and Future Work 204References 20511 UAV-enabled Smart Healthcare Application for Next-generation Wireless Networks 20911.1 Introduction 20911.1.1 Motivation 20911.2 Related Works 21111.3 System Model 21311.3.1 System Components 21411.3.2 Cost Model 21511.3.3 Revenue Model 22111.3.4 Objective Function 22211.4 Federated Deep Reinforcement Learning 22311.4.1 Deep Reinforcement Learning 22311.4.2 Federated Advantage Actor Critic 22611.4.3 Federated Proximal Policy Optimization 22711.4.4 Proposed Solution 22711.5 A Direction for Performance Study 23011.6 Conclusion 23211.6.1 Challenges 23311.6.2 Future Work 234References 23512 A Road Map for Personalized Medicine: Challenges and Innovations 23912.1 Introduction 23912.2 Use of Genome Data to Develop Personalized Medicine 24312.2.1 Genetic Biomarkers 24412.2.2 Biochemical Biomarkers 24412.3 Use of Medical Imaging Data to Develop Personalized Medicine 24512.4 Use of AI and ML in Drug Development in Personalized Medicine 24712.4.1 AI Applications in Pharmacogenomics 24812.4.2 Use of ML Models for Predicting Drug Responses 24912.4.3 Use of Deep Learning Models for Analyzing Genomic Data 25012.4.4 Use of AI and ML in Computational Modeling in Personalized Medicine 25012.5 Use of Digital Twin for Personalized Medicine 25112.6 Current Challenges and Future Directions 253References 25513 Delay-sensitive, Privacy-preserving Blockchain-enabled Fog-assisted Framework for Smart Healthcare 26313.1 Introduction 26313.2 Related Work 26513.3 System Model 26713.3.1 Founding Phase 26813.3.2 Bidding Phase 26913.3.3 Calculation 27013.4 Problem Formulation 27113.4.1 Criticality of the Data 27113.4.2 Data Transmission Cost 27213.4.3 Reputation Mechanism 27213.4.4 Profit of the Miners 27613.4.5 Profit Calculation for Patients 27713.4.6 Profit Calculation for Companies 27713.5 Proposed Solution 27813.5.1 Genetic Algorithm 27813.5.2 Problem Representation 28013.5.3 Problem Encoding 28113.5.4 Fitness Function 28213.5.5 Stopping Criteria 28313.5.6 Implementation Details 28313.6 Experimental Results 28613.6.1 Dataset 28613.6.2 Test and Results 28813.7 Conclusion 294Appendix 13.A 29413.a.1 NP-hard Problems 29413.a.2 0/1 Knapsack Problem 29413.a.3 NP-hard Proof 295References 296A Research Discussion, Tools, and Use Cases 299A. 1 Artificial Intelligence in Smart Healthcare 299A. 2 Research Trends in AI for Healthcare 300A. 3 Use Cases of AI in Smart Healthcare 300A. 4 Key Tools and Frameworks for AI-driven Healthcare 309A. 5 Challenges and Limitations in AI-driven Healthcare 312A. 6 Conclusion 313References 314Index 319