Deep Learning, Machine Learning and IoT in Biomedical and Health Informatics: Techniques and Applications

Cover
Half Title
Series Page
Title Page
Copyright Page
Table of Contents
Preface
Acknowledgement
Editors
Contributors
Part I: Machine Learning Techniques in Biomedical and Health Informatics
Chapter 1: Effect of Socio-economic and Environmental Factors on the Growth Rate of COVID-19 with an Overview of Speech Data for Its Early Diagnosis
1.1 Introduction
1.1.1 Motivation and Research Objective
1.2 Databases and Socioeconomic, Environmental Features
1.2.1 Temperature (f1)
1.2.2 Happiness Index (f2)
1.2.3 Cleanliness Index (f3)
1.2.4 Gross Domestic Product (f4)
1.2.5 Pollution Index (f5)
1.2.6 Number of Caregivers/Nurses per 1000 People (f6)
1.2.7 Number of Physicians per 1000 People (f7)
1.2.8 Diabetes Prevalence (f8)
1.2.9 Population Aged over Sixty-five (f9)
1.2.10 Smokers above Age Fifteen (f10)
1.3 Growth Rate Calculation and Feature Selection
1.3.1 Growth Rate Calculation
1.3.2 Feature Selection
1.4 COVID-19 Speech Analysis
1.5 Conclusion
References
Chapter 2: Machine Learning in Healthcare: The Big Picture
2.1 Overview of Machine Learning
2.1.1 Why Should Machines Learn?
2.1.2 What Is Machine Learning?
2.1.3 Types of Machine Learning
2.1.3.1 Deductive Learning
2.1.3.2 Inductive Learning
2.1.4 Deep Learning
2.2 Driving Forces for ML in Healthcare
2.2.1 Big Data in Healthcare
2.2.2 Demographic Shift
2.2.3 Global Pandemic
2.2.4 Pervasive Medical Errors
2.2.5 Patient-Centric Healthcare
2.3 Opportunities for ML Applications
2.3.1 Disease Diagnosis
2.3.2 Medical Imaging Analysis
2.3.3 Medical Prognosis
2.3.4 Smart Record and Personalized Medicine
2.3.5 Robotic Surgery
2.3.6 Genomics and Proteomics
2.3.7 Drug Discovery
2.3.8 Clinical Trial
2.3.9 Epidemic Outbreak and Control
2.4 Key Challenges for ML in Healthcare
2.4.1 Causality Problem
2.4.2 Data Limitations
2.4.3 Model Interpretability Problem
2.4.4 Adoption Barriers
2.5 Conclusions
References
Chapter 3: Heart Disease Assessment Using Advanced Machine Learning Techniques
3.1 Introduction
3.2 Literature Survey
3.3 Methodology
3.3.1 Proposed Methods
3.3.1.1 K-Nearest Neighbors (KNN)
3.3.1.2 Random Forest Classification
3.3.1.3 Support Vector Machines
3.3.1.4 Naive Bayes Classification
3.3.1.5 Logistic Regression
3.4 Results
3.5 Conclusion
References
Chapter 4: Classification of Pima Indian Diabetes Dataset Using Support Vector Machine with Polynomial Kernel
4.1 Introduction
4.2 Background Details
4.3 Dataset Description
4.4 Methodology
4.4.1 SVM (Support Vector Machine)
4.4.2 SVM Kernel
4.4.2.1 Polynomial Kernel Function
4.4.2.1.1 Homogenous Polynomial Kernel Function
4.4.2.1.2 Inhomogeneous Polynomial Kernel Function
4.4.2.2 Gaussian RBF Kernel Function
4.4.2.3 Sigmoid Kernel Function
4.4.2.4 Linear Kernel Function
4.5 Performance Measures
4.5.1 Classification Accuracy
4.5.2 Sensitivity
4.5.3 Specificity
4.6 Simulation and Experimental Result
4.6.1 Gain Graph
4.6.2 Response Graph
4.7 Conclusion
4.8 Future Scope
Reference
Chapter 5: Analysis and Prediction of COVID-19 Pandemic
5.1 Introduction
5.2 Literature Survey
5.3 Proposed System
5.4 The Data from the World
5.5 Results
5.6 Discussion
5.7 Conclusion
References
Chapter 6: Variational Mode Decomposition Based Automated Diagnosis Method for Epilepsy Using EEG Signals
6.1 Introduction: Background and Driving Forces
6.2 Literature Review
6.2.1 EEG Signal Processing Techniques
6.2.2 Models and Methods of Classification Used
6.3 Dataset
6.4 Variational Mode Decomposition
6.5 Features
6.5.1 Second and Fourth Order Difference Plot and Computation of Ellipse Area
6.6 Features: Renyi Entropy
6.7 Features Used: Average Amplitude
6.8 Multilayer Perceptron (MLP) Based Classification
6.9 Results and Discussion
6.10 Conclusion
References
Chapter 7: Soft-computing Approach in Clinical Decision Support Systems
7.1 Introduction
7.2 Literature Review
7.3 Incorporation of Databases into the CDSS System to Make It More Useful
7.4 CDSS Alerts, ADEs, and Physician Burnout
7.5 CDSS Security
7.6 Conclusion
7.6.1 Research Methodology
7.6.2 Findings
Bibliography
Chapter 8: A Comparative Performance Assessment of a Set of Adaptive Median Filters for Eliminating Noise from Medical Images
8.1 Introduction
8.2 Proposed Modified Circular Adaptive Median Filter
8.3 Simulation Study
8.3.1 Performance Indices
8.3.2 Simulation Results
8.4 Conclusion
Acknowledgment
References
Chapter 9: Early Prediction of Parkinson’s Disease Using Motor, Non-Motor Features and Machine Learning Techniques
9.1 Introduction
9.2 Review of Related Literature
9.3 Materials and Methods
9.3.1 Dataset Collection
9.3.2 Feature Descriptions
9.3.3 Data Pre-Processing
9.4 Model Description
9.5 Result and Discussion
9.6 Conclusion
9.7 Future Work
References
Part II: Deep Learning Techniques in Biomedical and Health Informatics
Chapter 10: Deep Neural Network for Parkinson Disease Prediction Using SPECT Image
10.1 Introduction
10.2 Methodology
10.2.1 Database
10.2.2 Image Processing
10.2.3 Methodology
10.2.3.1 Convolutional Neural Networks (CNN)
10.2.3.2 Results and Discussion
10.2.3.3 Comparison with Related Work
10.3 Discussion
10.4 Conclusion
References
Chapter 11: An Insight into Applications of Deep Learning in Bioinformatics
11.1 Introduction
11.2 Models in Deep Learning
11.2.1 Convolutional Neural Networks (CNN)
11.2.2 Recurrent Neural Network (RNN)
11.2.3 Autoencoder
11.2.4 Deep Belief Network (DBN)
11.3 Deep Learning in Bioinformatics
11.3.1 Deep Learning for Omics
11.3.2 Deep Learning for Biomedical Imaging
11.3.3 Deep Learning for Biomedical Signal Processing
11.3.4 Transfer Learning for Bio Informatics (TL for Bioinformatics)
11.3.5 Deep Reinforcement Learning for Bioinformatics
11.3.6 Deep Few Shot Learning for Bioinformatics
11.3.7 Deep Learning for Public Health
11.4 Deep Learning in Bioinformatics: Challenges and Limitations
11.5 Conclusion
References
Chapter 12: Classification of Schizophrenia Associated Proteins Using Amino Acid Descriptors and Deep Neural Network
12.1 Introduction
12.2 Protein Dataset Preparation
12.2.1 Protein Sequence Databases
12.2.2 STRING Network Analysis
12.2.3 Three Dimensional Structure of DRD2
12.3 Feature Table Generation
12.3.1 Amino Acid Composition
12.3.2 Physicochemical Properties
12.3.3 Composition Transition Distribution
12.3.4 FASGAI Vectors
12.4 Deep Neural Network
12.5 Conclusion
References
Chapter 13: Deep Learning Architectures, Libraries and Frameworks in Healthcare
13.1 Introduction
13.2 Deep Learning
13.2.1 Overview of Deep Learning
13.2.2 Deep Learning Frameworks and Libraries
13.2.2.1 Tensorflow
13.2.2.2 Keras
13.2.2.3 PyTorch
13.2.2.4 Caffe
13.2.2.5 MXNet
13.2.2.6 Chainer
13.2.2.7 Deeplearning4J
13.3 Basic Deep Learning Architectures
13.3.1 Convolutional Neural Networks
13.3.2 Recurrent Neural Networks
13.3.2.1 Long-Short-Term-Memory
13.3.2.2 Gated Recurrent Unit
13.3.3 Deep Belief Networks
13.3.4 Generative Adversarial Networks
13.3.5 Multilayer Perceptron
13.3.6 Fully Connected Neural Networks
13.4 Advanced Deep Learning Architectures
13.4.1 AlexNet
13.4.2 VCG Net
13.4.3 GoogLeNet
13.4.3.1 ResNet
13.4.4 Deep Recurrent CNN
13.4.5 Mask Scoring R-CNN
13.4.6 Ordered Neurons LSTM
13.4.7 Spherical CNNs
13.4.8 ResNeXt
13.4.9 YOLO
13.4.10 SegNet
13.4.11 SqueezeNet
13.5 Conclusion
References
Chapter 14: Designing Low-Cost and Easy-to-Access Skin Cancer Detector Using Neural Network Followed by Deep Learning
14.1 Introduction
14.1.1 Local and Offline Deployment
14.1.2 Eliminating Custom Hardware Requirements
14.1.3 Diversifying Classification
14.2 Computer-Aided System for Skin Cancer Diagnosis
14.3 Proposed Method
14.3.1 Flow of a CNN Model
14.3.2 Building CNN
14.3.2.1 Layer 1
14.3.2.2 Layer 2
14.3.2.3 Layer 3
14.3.2.4 Layer 4
14.3.2.5 Layer 5
14.3.3 Feature Extraction
14.3.4 Activation Functions
14.4 Results and Discussions
14.4.1 Comparison with Machine Learning Approach
14.4.1.1 Mean
14.4.1.2 Area
14.4.1.3 Border
14.4.2 Comparison with Other CNN Models
14.5 Future Scope
14.6 Conclusion
References
Part III: Internet of Things (IoT) in Biomedical and Health Informatics
Chapter 15: Application of Artificial Intelligence in IoT-Based Healthcare Systems
15.1 Introduction
15.2 Fuzzy Logic and Fuzzy Models of Health Care
15.3 Evolutionary Computing of Health Care
15.4 Artificial Neural Network for Health Care
15.5 A Probabilistic Model for Health Care
15.5.1 Risks of Probability Model in Healthcare
15.6 Big Data in Healthcare
15.6.1 Applications of Big Data in the Healthcare Sector
15.7 Data Mining in Healthcare
15.8 CI Applications in Healthcare
15.8.1 Increases Patient Engagement and Satisfaction
15.9 Organization of Deep Learning Applications for IoT in Healthcare
15.9.1 Internet of Healthy Things
15.9.2 Medical Diagnosis and Differentiation Applications
15.9.2.1 Automatic Diagnosis of Heart Disease
15.9.2.2 Automated EEG Disease Diagnosis
15.9.2.3 Cerebral Vascular Accidents (CVA) Diagnosis [ 60 ]
15.9.2.4 Detection of Atrial Fibrillation (AF)
15.9.2.5 Syndrome Differentiation
15.9.2.6 Diagnosis and the Treatment for Lung Cancer
15.9.2.7 Classifying Melanoma Diseases
15.10 Home-Based and Personal Healthcare
15.10.1 Disease Prediction Applications
15.10.2 IoMT Monitoring Solutions
15.11 Medical Internet of Things
15.11.1 Analysis of the Physiological Parameters
15.12 Rehabilitation Systems
15.13 Skin Pathologies and Dietary Assessment
15.14 Epidemic Diseases Treatment and Location-Aware Solutions
References
Chapter 16: Computational Intelligence in IoT Healthcare
16.1 Introduction
16.2 Edge Intelligence in Healthcare System
16.3 Smart Healthcare Delivery System
16.4 AI on Edge Architecture in Computational Intelligence for Healthcare System
16.5 Role of Artificial Intelligence in Diabetes Mellitus Management
16.6 Role of AI in Cardiovascular Disease Management
16.7 Role of AI in Neurodegenerative Diseases
16.8 Challenges of Computational Intelligence in IoT Healthcare
16.9 Role of AI in Helicobacter Pylori Detection
16.10 Conclusion and Future Perspectives
References
Chapter 17: Machine Learning Techniques for High-Performance Computing for IoT Applications in Healthcare
17.1 Introduction
17.2 The application of IoT in the Healthcare System
17.3 Data in Machine Learning for Healthcare
17.4 Traditional Centralized Learning: Machine Learning Runs in the Cloud, Gathering Data from Different Hospitals
17.5 Machine Learning Applications in Disease Prediction
17.5.1 Cancer
17.5.2 Diabetes
17.5.3 Cardiovascular Diseases
17.5.4 Chronic Kidney Disease
17.5.5 Parkinson Disease
17.5.6 Dermatological Diseases
17.6 Issues and Challenges
17.7 Conclusions
References
Chapter 18: Early Hypertensive Retinopathy Detection Using Improved Clustering Algorithm and Raspberry PI
18.1 Introduction
18.2 Preliminaries
18.2.1 Particle Swarm Optimization Clustering
18.2.2 Raspberry PI
18.3 Related Work
18.4 Methodology
18.4.1 Pre-processing
18.4.2 Segmentation
18.4.2.1 Elevated Continuous Particle Swarm Optimization Clustering
18.4.2.2 Feature Extraction
18.5 Experimental Results
18.5.1 Performance Analysis
18.6 Conclusion
References
Chapter 19: IoT Based Elderly Patient Care System Architecture
19.1 Introduction
19.2 Healthcare System Without Patient Mobility Support
19.3 Healthcare System with Patient Mobility Support
19.4 Existing Architectures of IoT Based Health Care System
19.4.1 Comparison of the Existing Health Care Systems
19.5 Proposed Architecture of IoT Based Elderly Patient Care System
19.5.1 Features of the Proposed Architecture
19.5.2 Advantages of the Proposed System
19.6 Discussion
19.7 Conclusion
References
Index