Blockchain and Machine Learning for e-Healthcare Systems

✍ Scribed by Balusamy Balamurugan (editor), Naveen Chilamkurti (editor), T. Lucia Agnes Beena (editor), T. Poongodi (editor)

• Publisher: Institution of Engineering and Technology
• Year: 2021
• Tongue: English
• Leaves: 481
• Series: Healthcare Technologies
• Category: Library

✦ Synopsis

Blockchain and machine learning technologies can mitigate healthcare issues such as slow access to medical data, poor system interoperability, lack of patient agency, and data quality and quantity for medical research. Blockchain technology facilitates and secures the storage of information in such a way that doctors can see a patient's entire medical history, but researchers see only statistical data instead of any personal information. Machine learning can make use of this data to notice patterns and give accurate predictions, providing more support for the patients and also in research related fields where there is a need for accurate data to predict credible results.

This book examines the application of blockchain technology and machine learning algorithms in various healthcare settings, covering the basics of the technologies and exploring how they can be used to improve clinical outcomes and improving the patient's experience. These topics are illustrated with reference to issues around the supply chain, drug verification, reimbursement, control access and clinical trials. Case studies are given for applications in the analysis of breast cancer, hepatitis C, and COVID-19.

✦ Table of Contents

Cover
Contents
About the editors
Preface
1 Blockchain technology and its relevance in healthcare
1.1 Introduction
1.1.1 Evolution of blockchain technology
1.1.2 Characteristics of blockchain technology
1.1.3 Overview of blockchain architecture
1.1.4 Merkle tree structure
1.2 Basic components of blockchain
1.2.1 Cryptographic hash functions
1.2.2 Asymmetric-key cryptography
1.2.3 Transactions
1.2.4 Ledgers
1.2.5 Blocks
1.3 Consensus models
1.3.1 Proof of work
1.3.2 Proof of stake
1.3.3 Practical byzantine fault tolerance
1.3.4 Delegated proof of stake
1.3.5 Round robin consensus model
1.3.6 Proof of authority (identity) model
1.3.7 Proof of elapsed time (PoET) consensus model
1.4 Challenges and opportunities of blockchain technology
1.4.1 Security and privacy of the data
1.4.2 Storage
1.4.3 Standardization
1.4.4 Scalability
1.4.5 Interoperability
1.4.6 Key management
1.4.7 Blockchain vulnerabilities
1.4.8 Social challenges
1.4.9 Accountability
1.4.10 Accuracy
1.4.11 Agility
1.4.12 Fighting counterfeit drugs
1.4.13 Cost efficient
1.4.14 Improving research and development
1.5 Types of blockchain
1.5.1 Public blockchain
1.5.2 Private blockchain
1.5.3 Consortium blockchain
1.5.4 Permissioned blockchain
1.5.5 Permissionless blockchain
1.6 Relevance of blockchain for healthcare
1.6.1 Blockchain for medical record management
1.6.2 Blockchain for medicinal research
1.6.3 Blockchain for insurance claims
1.6.4 Blockchain for counterfeit drugs
1.6.5 Blockchain to prevent future pandemics
1.6.6 Blockchain to save cost
1.7 Conclusion
References
2 Privacy issues in blockchain
2.1 National and Corporate Support
2.2 Asia trade and European trade
2.3 Multinational policies vs blockchain
2.3.1 Security of blockchain?
2.3.2 Shifting security to the end user
2.3.3 Trade-offs
2.3.4 Key developments of blockchain for voting
2.3.5 Improving productivity in agriculture
2.3.6 Guarantee straightforwardness, supportability in fishing
2.3.7 Cryptocurrency regulations
2.3.8 Energy industry
2.4 Blockchain approaches to data privacy in healthcare
2.4.1 Blockchain for electronic medical record (EMR) data management
2.4.1.1 Blockchain and health-care data protection
2.4.2 Blockchain for personal health record (PHR) data management
2.4.3 Blockchain for point-of-care genomics
2.4.4 Blockchain for EHR data management
2.4.5 Fast health-care interoperability resources
2.4.6 Health-care blockchain
2.4.7 On-chain
2.4.7.1 High-level data
2.4.8 Off-chain
2.4.8.1 Large data files
2.4.8.2 Large data files
2.4.9 Network is the concern not a database
2.4.10 Clear definition of use cases
2.4.11 Throughput and scalability
2.4.12 Adequate data
2.4.13 Blockchain privacy poisoning
2.4.14 Consent management and the blockchain
2.5 Blockchain privacy poisoning in the context of other privacy issues
2.5.1 Who should be accountable for blockchain privacy poisoning?
2.5.2 Problems of blockchain security/privacy
2.5.2.1 MITM attacks
2.5.2.2 Data tampering
2.5.2.3 DDoS attacks
2.5.2.4 Privacy
2.5.3 Challenges
2.6 Blockchain security for health data: promises, risks, and future development: blockchain security issues
2.6.1 Challenges and limitations
2.6.1.1 Data ownership and privacy
2.6.1.2 Legal
2.6.1.3 Security
2.6.1.4 Future of blockchain
2.7 Conclusion
References
3 Reforming the traditional business network
3.1 Introduction
3.2 Applications of blockchain
3.2.1 Blockchains in electronic health records (EHR)
3.2.2 Blockchains in clinical research
3.2.3 Blockchains in medical fraud detection
3.2.4 Blockchains in neuroscience
3.2.5 Blockchains in pharmaceutical industry and research
3.3 Business benefits of blockchain
3.4 Reliance on blockchain usage
3.4.1 Protection claims
3.4.2 Gold supply chain
3.4.3 Coordination’s activities
3.5 Market resistance to blockchain
3.5.1 Resistance
3.5.2 Level support and resistance
3.5.3 Polarity
3.6 Role of blockchain in healthcare
3.6.1 Drug supply chain
3.6.2 Clinical data exchange and interoperability
3.6.3 Billing and claims management
3.6.4 Cybersecurity and healthcare IoT
3.6.5 Population health research and pharma clinical trials
3.7 Blockchain in hospital management services
3.7.1 Blockchain in healthcare
3.7.2 New business opportunities
3.7.3 Electronic medical records
3.7.4 Guideline compliance
3.7.5 Decreased billing and speedy claim settlement
3.7.6 Decrease in information risks
3.7.7 Coordination of data
3.8 Blockchain—the new age business disruptor
3.8.1 3D printing
3.8.2 Accounting
3.8.3 Agriculture
3.8.4 Art
3.8.5 Credit management
3.8.6 Compliance
3.8.7 The Internet of Things (IoT)—connected devices
3.9 Conclusion
References
4 A deep dive into Hyperledger
4.1 Hyperledger Frameworks
4.1.1 Hyperledger Besu
4.1.2 Hyperledger Burrow
4.1.3 Hyperledger Fabric
4.1.4 Hyperledger Indy
4.1.5 Hyperledger Iroha
4.1.6 Hyperledger Sawtooth
4.1.7 Hyperledger Grid
Summary of Hyperledger frameworks
4.2 Hyperledger Libraries
4.2.1 Hyperledger Aries
4.2.2 Hyperledger Quilt
4.2.3 Hyperledger Transact
4.2.4 Hyperledger Ursa
4.3 Hyperledger Tools
4.3.1 Hyperledger Avalon
4.3.2 Hyperledger Caliper
4.3.3 Hyperledger Cello
4.3.4 Hyperledger Explorer
4.3.5 Hyperledger Composer
4.4 Blockchain in enterprise
4.4.1 Use cases
4.4.1.1 Healthcare
4.4.1.2 Banking
4.4.1.3 Supply chain management
4.4.1.4 Agriculture
4.4.1.5 Others
4.5 Blockchain in e-healthcare
4.5.1 Improve medical record access
4.5.2 Improve clinical trials
4.5.3 Improve drug traceability
Healthcare areas where blockchain can be applied
4.6 An example of healthcare data management using IBM blockchain platform
References
5 Machine learning
5.1 Introduction
5.1.1 Machine learning life cycle
Step 1: Collecting data
Step 2: Data preprocessing
Step 3: Data cleaning
Step 4: Model building
Step 5: Training model
Step 6: Testing model
Step 7: Implementation
5.1.2 Components in machine learning
5.2 Different types of learning
5.2.1 Supervised learning
5.2.1.1 Types of supervised machine learning algorithms
5.2.1.2 Difference between regression and classification
5.2.1.3 Challenges in supervised machine learning
5.2.1.4 Advantages of supervised learning
5.2.1.5 Disadvantages of supervised learning
5.2.2 Unsupervised learning
5.2.2.1 Types of unsupervised learning algorithm
5.2.2.2 Applications of unsupervised machine learning
5.2.2.3 Disadvantages of unsupervised learning
5.2.2.4 Difference between supervised learning and unsupervised learning
5.2.3 Reinforcement learning
5.2.3.1 How reinforcement learning works?
5.2.3.2 Characteristics of reinforcement learning
5.2.3.3 Reinforcement learning vs. supervised learning
5.2.3.4 Applications of reinforcement learning
5.2.3.5 Challenges of reinforcement learning
5.3 Types of machine learning algorithms
5.3.1 Classification algorithms
5.3.1.1 Naı¨ve Bayes classification
5.3.1.2 Nearest neighbor
5.3.1.3 Support vector machine
5.3.1.4 Decision trees
5.3.1.5 Neural networks
5.3.2 Regression algorithm
5.3.2.1 Linear regression
5.3.2.2 Logistic regression
5.3.3 Dimensionality reduction algorithm
5.3.3.1 Principal component analysis
5.3.3.2 Radical basis function
5.3.4 Clustering algorithms
5.3.4.1 K-Means algorithm
5.3.4.2 DB scan algorithm
5.3.4.3 Gaussian mixture model
5.3.4.4 EM algorithm
5.3.5 Reinforcement algorithm
5.3.5.1 Deep Q networks
5.3.5.2 Deep deterministic policy gradient
5.3.5.3 Asynchronous advantage actor critic
5.3.6 Machine learning in healthcare
5.3.6.1 Identification of diseases and diagnosis
5.3.6.2 Drug discovery and manufacturing
5.3.6.3 Medical imaging
5.3.6.4 Personalized medicine/treatment
5.3.6.5 Smart health records
5.3.6.6 Predicting diseases
5.3.7 Advantages and disadvantages of machine learning
5.3.7.1 Advantages of machine learning
5.3.7.2 Disadvantages of machine learning
5.3.8 Limitations of ML in healthcare industry
5.4 Conclusion
References
6 Machine learning in blockchain
6.1 Introduction
6.1.1 What is machine learning?
6.1.2 Importance of ML in blockchain
6.1.3 Merits and demerits
6.2 Types of ML
6.2.1 Supervised learning
6.2.2 Unsupervised learning
6.2.3 Reinforcement learning
6.3 Different ML algorithms
6.3.1 Linear regression
6.3.2 Logistic regression
6.3.3 Decision tree and SVM
6.3.4 Naı¨ve Bayes
6.3.5 k-Nearest neighbor
6.3.6 K-Means
6.3.7 Gradient boosting algorithms—GBM, XGBoost, LightGBM, CatBoost
6.4 Significance of ML in the health-care industry
6.4.1 Identifying diseases and diagnosis
6.4.2 Drug discovery and manufacturing
6.4.3 Medical imaging diagnosis
6.4.4 Personalized medicine
6.4.5 Machine-learning-based behavioral modification
6.4.6 Smart health records
6.4.7 Clinical trial and research
6.4.8 Crowd-sourced data collection, better radiotherapy and outbreak prediction
6.5 Implementation difficulties of using ML in healthcare
6.6 Applications and future scope of research
6.7 Conclusion
References
7 Framework for approaching blockchain in healthcare using machine learning
7.1 Introduction
7.1.1 Introduction to machine learning
7.1.2 Introduction to blockchain
7.2 The steps in machine learning
7.3 Gathering health data
7.3.1 Influence of data assemblage in healthcare
7.3.2 Recent trends in data collection
7.3.3 Healthcare datasets
7.4 Data preparation
7.4.1 Benefits of data preparation and the cloud
7.4.2 Data preparation steps
7.5 Choosing a model
7.5.1 Types of machine learning algorithms
7.5.1.1 Supervised learning
7.5.1.2 Unsupervised learning
7.5.1.3 Reinforcement learning
7.5.2 Most familiar machine learning algorithms
7.5.2.1 Linear regression
7.5.2.2 Logistic regression
7.5.2.3 Decision tree
7.5.2.4 Support vector machine
7.5.2.5 Naïve Bayes
7.5.2.6 k-Nearest neighbours
7.5.2.6 k-Nearest neighbours
7.5.2.7 K-Means
7.5.2.8 Random forest
7.5.2.9 Dimensionality reduction algorithms
7.5.2.10 Gradient boosting algorithms
7.5.2.11 XGBoost
7.5.2.12 LightGBM
7.5.2.13 CatBoost
7.5.3 Need for models in healthcare using blockchain
7.5.3.1 Description of healthcare interoperability
7.6 Training
7.6.1 The purpose of train/test sets
7.6.2 Blockchain for privacy in healthcare
7.6.3 Quantum of training data requirements
7.7 Evaluation
7.7.1 Evaluation metrics
7.7.2 Evaluation metrics and assessment of machine learning algorithms in healthcare
7.8 Parameter tuning
7.9 Predictive analytics
7.9.1 Requirement collection
7.9.2 Data collection
7.9.3 Data analysis and massaging
7.9.4 Statistics, machine learning
7.9.5 Predictive modelling
7.9.6 Prediction and monitoring
7.10 Benefits of integrating machine learning and blockchain
7.11 Conclusion
References
8 Reforming the traditional business network
8.1 Introduction
8.2 Artificial intelligence in healthcare
8.2.1 Artificial intelligence doctors
8.2.2 AI—robot treatment
8.2.3 AR/VR treatment
8.2.4 Non-adherence to prescriptions
8.3 Blockchain in healthcare
8.3.1 Blockchain in healthcare
8.3.2 Medical credential tracking
8.3.3 Drug trials
8.4 Linear algebra in ML
8.4.1 Dataset and data files
8.4.2 Images and photographs
8.4.3 One-hot encoding
8.4.4 Applications
8.4.4.1 Clinical trial design
8.4.4.2 Medical imagery
8.4.4.3 Treatment planning for nuclear medicine
8.5 New medical imaging modalities
8.5.1 Multivalued data images
8.5.2 Phase contrast magnetic resonance angiography (MRA)
8.5.3 Diffusion tensor MRI
8.5.4 Federated tensor factorization
8.6 Medical appliance of norms
8.6.1 Significance of medical devices
8.6.2 Medical device safety
8.6.3 Global Harmonization Task Force
8.6.4 Classification of medical devices
References
9 Healthcare analytics
9.1 Introduction
9.2 Analytics
9.2.1 Descriptive analytics
9.2.2 Predictive analytics
9.2.3 Perspective analysis
9.3 Emerging technologies in healthcare analytics
9.3.1 Big data technology in healthcare analytics
9.3.2 Internet of Things in healthcare analytics
9.3.2.1 IoT for patients
9.3.2.2 IoT for physicians
9.3.2.3 IoT for hospitals
9.3.2.4 IoT for insurance companies
9.3.3 Artificial intelligence in healthcare
9.3.4 Blockchain in healthcare
9.4 History of healthcare analytics
9.5 Exploring software for healthcare analytics
9.5.1 Anaconda
9.5.2 SQLite
9.6 Challenges with healthcare analytics
9.6.1 High-dimensional data
9.6.2 Irregularities in data
9.6.3 Missing data
9.7 Conclusion
References
10 Blockchain for healthcare
10.1 Introduction
10.1.1 Bitcoin blockchain
10.1.2 Block
10.1.3 Chain
10.2 Ethereum blockchain
10.3 Contracts and healthcare: the arising need of smart contracts
10.3.1 Smart contracts
10.3.2 Zero-knowledge-proofs and smart contracts
10.3.3 Ricardian contracts for healthcare
10.3.4 Hybrid smart–Ricardian contracts
10.4 Applications of blockchain
10.5 Healthcare data
10.5.1 Structured data sets
10.5.2 Non-structured data sets
10.6 Popular resources for gathering healthcare data
10.7 Need of healthcare data
10.8 Services offered by the blockchain in healthcare
10.8.1 Data sharing and privacy issues
10.8.2 Longitudinal patient records and health data accuracy
10.8.3 Drug track ability
10.8.4 Fake medical credentials
10.8.5 Claims processing
10.8.6 Supply chain management
10.8.7 Interoperability of data among medical institutes
10.9 Medicines and supply chain tracking enabled by blockchain
10.10 Data security concerns in EMR and healthcare domain
10.11 Choices of blockchain platforms for healthcare
10.11.1 Ethereum
10.11.2 IBM® blockchain
10.11.3 Hyperledger
10.11.4 Hydrachain
10.11.5 R3 Corda
10.11.6 MultiChain
10.11.7 BigchainDB
10.11.8 OpenChain
10.11.9 Quorum blockchain platform
10.11.10 EOS blockchain platform
10.11.11 Internet-of-Things application (IOTA) blockchain platform
10.12 Major healthcare blockchain use cases under development
10.13 Storage challenges and need for inter planetary file system (IPFS) enabled blockchain for healthcare
10.14 IPFS
10.15 Why do we need IPFS?
10.16 Blockchain and IPFS
10.17 Challenges and roadblocks to the realisation of blockchain-enabled healthcare
10.18 Conclusion
References
11 Improved interop blockchain applications for e-healthcare systems
11.1 Introduction
11.1.1 How blockchain is used in healthcare?
11.1.2 Challenges in interoperability between various sections of healthcare system
11.2 Literature review
11.2.1 Electronic health records
11.2.2 Drug tracking
11.2.3 Blockchain in future healthcare
11.2.4 Blockchain and cryptocurrencies
11.2.4.1 Ethereum
11.2.4.2 Smart contracts
11.2.4.3 Ethereum virtual machine and DApps
11.2.4.4 SWARM
11.3 Proposed method
11.3.1 EHR—architecture
11.3.2 Smart-contract system design
11.3.3 Smart contract implementation
11.3.4 Drug traceability using blockchain
11.3.4.1 Why blockchain?
11.3.4.2 What is supply chain?
11.3.4.3 Supply chain implementation
11.3.4.4 Entering data in the system
11.3.4.5 Roles and data access
11.3.5 Clinical trials using blockchain
11.3.5.1 What are clinical trials?
11.3.5.2 How blockchain is useful?
11.3.5.3 Methods for implementation of the clinical trials
11.4 Conclusion and future scope
References
12 Blockchain: lifeline care for breast cancer patients in developing countries
12.1 Introduction
12.2 Blockchain
12.2.1 Key attributes
12.2.2 Kind of blockchains
12.2.3 Agreement instruments
12.2.4 Shrewd agreements
12.2.5 The capability of blockchain in the human services space
12.3 Healthcare data management
12.3.1 Blockchain-based smart contracts for healthcare
12.3.2 The process for issuing and filling of medical prescriptions
12.3.3 Sharing laboratory test/results data
12.3.4 Enabling effective communication between patients and service providers
12.3.5 Smart-contracts-based clinical trials
12.4 Healthcare data management
12.4.1 Medication revelation and pharmaceutical research
12.4.2 Flexibly chain and counterfeit medications discovery
12.5 Challenges and future scope
12.5.1 Interoperability and integration with the legacy systems
12.5.2 Selection and motivating forces for support
12.5.3 Uncertain expense of activity
12.5.4 Regulation
12.5.5 Governance
12.5.6 Scaling
12.6 Conclusion
References
13 Machine learning for health care
13.1 Machine learning pipelining
13.2 Applications of ML in health care
13.3 Common machine learning approaches in machine learning
13.3.1 Artificial neural networks
13.3.2 Tree-like reasoning
13.3.3 Other common ML algorithms
13.4 Application of machine learning in health care
13.4.1 COVID-19—interpretation, detection and drug discovery using machine learning
13.4.1.1 Interpretation of COVID-19
13.4.1.2 Detection approaches for COVID-19 using ML algorithms
13.4.1.3 Drug discovery for COVID-19
13.4.1.4 Machine learning for chronic disease analysis
13.5 Breaking the blackbox of neural networks through explainable AI
13.6 Conclusion
References
14 Machine learning in healthcare diagnosis
14.1 Introduction
14.1.1 State of art of diagnosing system using machine learning
14.2 Heart diseases diagnosing system using machine learning
14.2.1 Various methods for diagnosis of heart disease using ML
14.2.1.1 Naı¨ve Bayes classifier
14.2.1.2 Random forest
14.2.1.3 Simple logistic regression
14.2.1.4 Artificial neural networks
14.2.1.5 Support vector machine
14.2.1.6 Architecture flow for heart disease using ML
14.3 Breast cancer diagnosing system using machine learning
14.3.1 k-NN method for breast cancer prediction
14.4 Neurological diseases diagnosing system using machine learning
14.4.1 Detecting the neurodegenerative diseases and the traumatic brain related injuries using D-CNN
14.4.2 Diagnosis using 3D-CNN
14.4.3 Training of 3D sparse autoencoder and 3D-CNN
14.5 Challenges and future direction of medical diagnosing system
14.6 Conclusion
References
15 Python for healthcare analytics made simple
15.1 Introduction
15.1.1 Data
15.1.2 Importance of data quality in healthcare
15.1.3 Elements of data quality
15.1.4 Ensuring data and information quality
15.2 Data extraction
15.2.1 Implementing NER with NLTK
15.2.2 Implementing Named Entity Recognition using SpaCy
15.3 Data visualization and tools
15.4 Advanced visualization methods
15.5 Data analytics
15.5.1 Descriptive analytics
15.5.2 Diagnostic analytics
15.5.3 Predictive analytics
15.5.4 Prescriptive analytics
15.6 Healthcare and technology: open issues
15.6.1 Data remanence
15.6.2 Data interoperability
15.6.3 Data staging
15.6.4 Application of big data in biomedical research
15.7 Conclusion
References
16 Identification and classification of hepatitis C virus: an advance machine-learning-based approach
16.1 Introduction
16.2 Literature survey
16.2.1 Artificial neural network
16.2.2 Random forest
16.2.3 Decision tree
16.2.4 Support vector machine
16.3 Proposed methodology
16.3.1 Bagging classifier
16.4 Experimental setup
16.4.1 Data preprocessing
16.4.2 Dataset description
16.4.3 Attribute information
16.4.4 Evaluation metrics
16.4.4.1 Confusion matrix
16.4.4.2 Accuracy
16.4.4.3 Error
16.4.4.4 Precision
16.4.4.5 Recall
16.4.4.6 False positive rate
16.4.4.7 F1 score
16.4.5 Environmental setup
16.5 Result analysis
16.6 Conclusion
References
17 Data visualization using machine learning for efficient tracking of pandemic – COVID-19
17.1 Introduction
17.2 Data preprocessing
17.2.1 Importance of data preprocessing
17.2.1.1 Categorical
17.2.1.2 Numeric
17.2.2 Data preprocessing consist of following steps
17.2.2.1 Data clean-up
17.2.2.2 Aggregation
17.2.2.3 Sampling
17.2.2.4 Dimensionality reduction
17.2.2.5 Discretization
17.3 Exploratory data analysis
17.3.1 Univariate analysis
17.3.1.1 Measures of central tendency
17.3.1.2 Measures of dispersion
17.3.2 Bivariate analysis
17.4 Data visualization techniques
17.4.1 Box plot
17.4.2 Charts
17.4.2.1 Bar chart
17.4.2.2 Line chart
17.4.2.3 Pie chart
17.5 Maps
17.5.1 Thickness maps
17.5.2 Tree map
17.6 Disperse plot
17.7 Gantt chart
17.8 Importance of data visualization in healthcare
17.8.1 Uses of data visualization
17.9 COVID-19 gripping the world
17.10 COVID-19 India situation
17.11 Issue and challenges
17.12 Conclusion
References
Index
Back Cover