Explainable Artificial Intelligence in Healthcare Systems

✍ Scribed by A. Anitha Kamaraj

• Tongue: English
• Leaves: 389
• Edition: 1
• Category: Library

✦ Table of Contents

Contents
Preface
Acknowledgements
Section 1. Primitive Concepts under Explainable AI
Chapter 1
XAI in Healthcare: Black Box to Interpretable Models
Abstract
1.1 Introduction
1.2 Significance of XAI
1.3 Motivation for Applying XAI in Health Care
1.4 Advantages in Healthcare
1.5 XAI Methods
1.6 Commonly Used XAI Methods
1.7 Interprétable Machine Learning Model
1.8 Comparative Study of Four Models
1.9 Applications of XAI
1.10 Limitations of XAI and Future Developments and Trends
1.11 Conclusion
References
Chapter 2
Healthcare XAI: A Systematic Study
Abstract
2.1 Introduction
2.2 Framework from Machine Learning to XAI
Model-Specific vs Model-Agnostic
Global Methods vs Local Methods
Before-Modelling vs During-Modelling vs After-Modelling
Surrogate Techniques vs Visualization Techniques
2.3 Knowledge Representation of XAI
2.4 Interpretability and Explainability
Interpretability
Explainability
Importance of Interpretability and Explainability
Interpretability vs Explainability
2.5 XAI Approaches and Analysis in Sustainable Smart Healthcare Informatics
2.6 Role of AI
2.7 XAI Approaches
2.8 On Federated Learning and the Role of Federated Learning with XAI in Health Sector
2.9 Threats of XAI in Healthcare
2.10 Opportunities of XAI
2.11 Results and Analysis
2.12. Conclusion
References
Chapter 3
XAI in Healthcare: A SWOT Analysis
Abstract
3.1 Introduction
3.2 Healthcare 5.0
3.3 XAI
3.3.1. Definitions
3.4 XAI Process
3.5 Pillars of XAI
3.6 XAI – SWOT Analysis
3.6.1 Strength
3.6.2 Weakness
3.6.3 Opportunities
3.6.4 Threats
3.7. Conclusion
References
Chapter 4
A Comprehensive Review on the Developments in Explainable Artificial Intelligence
Abstract
4.1 Introduction
4.2 Explainable Artificial Intelligence Literature Review
4.3 Explainable AI Systems
4.4 Applications of XAI
4.4.1 XAI in Various Domains
4.4.2 XAI in Industrial Applications
4.5 Advancements in XAI
4.5.1 Logic-Based Strategy
4.5.2 Taxonomies
4.5.3 Real-Time Explanations
4.5.4 Using Graph Neural Networks in XAI
4.5.5 Wikipedia Knowledge Graph
4.5.6 Frameworks
4.5.7 Changes Suggested in XAI
4.6 Advantages and Disadvantages of XAI Systems
4.7 Conclusion and Future Scope
References
Chapter 5
Unveiling the Algorithms: How Explainable AI Reshapes Healthcare
Abstract
5.1 Introduction
5.2 Role of AI in Healthcare
5.3 The Importance of XAI in Healthcare
5.4 Interpretable vs. Explainable AI
5.5 The Impact of Black Box AI in Healthcare
5.6 Case Studies of XAI in Healthcare
5.7 XAI Techniques
5.7.1 User-Centered Design for Explainable AI
5.8 Advantages and Limitations of Explainable AI
5.9 Challenges and Opportunities of XAI in Healthcare
5.10 Real-World Use Cases of Explainable AI in Healthcare
5.11 Future Trends
5.12 Conclusion
References
Section 2. Explainable AI in Smart Telemedicine and Telehealth
Chapter 6
Sensor Scheduling in an IoT Health Monitoring System with Interference Awareness
Abstract
6.1 Introduction
6.2 Internet of Things
6.3 IoT in Healthcare
6.4 Create a Framework for Monitoring Health
Open-Source Hardware
Sensor
Connection for Application
6.5 Acquiring Information from Sensors
6.6 Suggested Method
6.7 Results
6.8 Conclusion
References
Chapter 7
A Time Series-Based Artificial Neural Networks for Predicting COVID-19 Positive Cases in Indonesia
Abstract
7.1 Introduction
7.2 Literature Review
7.3 Proposed Methodology
7.3.1 Data Collection
7.3.2 Data Normalization
7.4 Artificial Neural Networks (ANNs)
7.5 ANNs Training Algorithms
7.5.1 Evaluation Metric Performance
7.6 Experimental Results
7.6.1 Results of ANN-LM
7.6.2 Results of ANN-BFGS
7.6.3 Results of ANN-SCG
7.7 Conclusion
References
Chapter 8
Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope Using Deep Learning Algorithm
Abstract
8.1 Introduction
8.2 Types of Detection of Abnormal Sounds from Lungs
Auscultation
Pulmonary Function Tests (PFTs)
Imaging Techniques
Bronchoscope
Blood Tests
Summary
8.3 Literature Survey
8.3.1 Machine Learning (ML) Algorithms for Detecting Abnormal Sounds in Lungs
8.3.2 SVM for Abnormal Sound Detection in Lungs Using Vest Coat Stethoscope
8.3.3 Random Forest (RF) for Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope
8.3.4 Decision Tree for Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope
The Decision Tree Approach
Benefits of the Decision Tree Approach
8.3.5 K-Means Algorithm for Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope
Background
K-Means Algorithm
Application in Abnormal Sound Detection
Benefits and Challenges
8.3.6 Linear Regression for Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope
Linear Regression for Abnormal Sound Detection
Data Collection and Preprocessing
Model Training and Evaluation
Potential Challenges and Future Directions
8.4 Deep Learning Algorithms for Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope: An Introduction
8.5 Proposed AI-Powered Vest-Coat (AI-VC)
Dataset Description
8.6 CNN for Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope
Introduction
Vest-Coat Stethoscope
Convolutional Neural Network (CNN) for Abnormal Sound Detection
Benefits and Potential Applications
8.7 Auto Encoder for Abnormal Sound Detection in Lungs Using Vest-Coat Stethoscope
Limitations of Traditional Auscultation
Proposed Solution
Methodology
Significance of the Study
8.8 Performance Metrics
8.9 Conclusion
References
Section 3. Public Health Application using Explainable AI
Chapter 9
Heart Diseases Prediction Based on Multiple Machine Learning Models
Abstract
9.1 Introduction
9.1.1 Classification
9.2 Background Study
9.3 Generalised Prediction and Analysis
9.4 Methodology
9.4.1. Metrics to Evaluate Performance
9.5. Experimental Setup
9.5.1. Data Description
9.5.2. Data Preprocessing
9.5.3. Supervised Machine Learning Algorithms Classification Techniques
9.5.3.1. Logistic Regression
9.5.3.2. Support Vector Machine
9.5.3.3. Naïve Bayes
9.5.3.4. Random Forest
9.5.4. Analysis Using Logistic Regression (LR)
9.5.5. Analysis Using Support Vector Machine (SVM)
9.5.6. Analysis Using Naïve Bayes (NV)
9.5.7. Analysis Using Random Forest (RF)
9.6. Prediction Comparative and Result Analysis among Different Models
9.7 Conclusion and Future Scope
References
Chapter 10
Artificial Intelligence and Explainable Artificial Intelligence Applications for Predicting and Preventing Complications in Maternal Health
Abstract
10.1 Introduction
10.1.1. Overview of Pregnancy and Maternal Care
10.1.2. Artificial Intelligence and Machine Learning in Maternal Care
10.2. Applications of AI and ML in Pregnancy Risk Assessment
2.1. Estimation of Foetal Gestational Age
10.2.2. CRL Measurement Is Essential for Accurate Estimation of Gestational Age
10.2.3. Prediction of Foetal Growth Restriction
10.2.4. Detection of Congenital Anomalies
10.2.5. Prediction of Gestational Diabetes Mellitus
10.2.6. Prediction of Preeclampsia
10.2.7. Prediction of Preterm Labor
10.3. Applications of AI in Delivery and Postpartum
10.3.1. Real-Time Monitoring During Labor
10.3.2. Postpartum Haemorrhage
10.3.3. Breastfeeding Support
10.3.4. Postpartum Depression Prediction
10.4. Medication Safety
10.5. Clinical Decision System
10.6 Explainable Artificial Intelligence
10.7 Ethical and Legal Considerations
10.8 Challenges and Future Directions
10.9 Conclusion
References
Chapter 11
Heart Disease Prediction Using Machine Learning Algorithm with Explainable Artificial Intelligence for Health Care System
Abstract
11.1 Introduction
11.2 Literature Survey
11.2.1. Background Work
11.2.2. Logistic Regression
11.2.3. Random Forest
11.2.4. K-Nearest Neighbour
11.2.5. Explainable Artificial Intelligence (XAI) Method LIME
11.3. Proposed Methodology
11.3.1. Pre-Processing
11.3.2. Classification Techniques
11.3.3. Description of the Output
11.3.4. Implementation and Testing
11.3.4.1. Dataset Description
11.4. Results
11.5 Conclusion
References
Chapter 12
Ailment Prophecy Based on Symptoms Using Machine Learning
Abstract
12.1 Introduction
12.2 Literature Review
12.3 Proposed Methodology
Module 1: Data Preparation
Module 2: Building the Model Using Random Forest Classifier
Module 4: Naive Bayes Model Construction
Module 5: Support Vector Machine Model Construction
Inferences
12.4 Performance Analysis
12.5 Conclusion
Chapter 13
Predicting the Disease Outbreak Using Artificial Intelligence and Data Mining Techniques
Abstract
13.1 Introduction
13.2 Literature Review
13.3 Methodology
13.3.1 Association Rule Mining
Concept
Associated Parameters
Support
Confidence
Lift
13.3.2 FP Growth – Association Rule Mining Algorithm
13.3.3. Graph Node Classification Using Airline Routes
Concept
Creating the Graph
Graph Node Embedding Algorithm - node2vec Algorithm
Random Walks
Skip-Gram Model
Negative Sampling
Optimization
Embedding Extraction
13.3.4 Logistic Regression
13.4 Novelty of the Proposed Approach
13.5 Experimental Data, Implementation and Results
Experimental Data
Association Rule Mining
Graph Node Classification
Disease Dataset
Airline Routes
Airport IATA Code - Country
Country Geographical Coordinates
13.6 Implementation
FP Growth Algorithm-Association Rule Mining
Preprocessing
Applying the FP Growth Algorithm
Generating Predictions
Graph node Classification
Preprocessing
Graph Formation
Graph Node Embedding
Logistic Regression
13.7 Evaluation
Accuracy
Precision
13.8 Results
Association Rule Mining
Graph Node Classification
13.9 Discussion and Conclusion
References
Chapter 14
Explainable AI in Medical Image Processing for Health Care
Abstract
14.1 Introduction
14.2 Explainable AI in Medical Image Processing
14.2.1 Major Challenges in XAI
14.2.2 Benefits of Explainable AI in Medical Image Processing
14.3 Classification Models
Decision Tree
Naive Bayes
Random Forest Classifier Model
A Support Vector Machine (SVM)
14.4 Literature Review
14.5 Exploratory Data Analysis
Data Cleaning
14.6 Explainable AI Techniques in Medical Image Processing
14.7 Medical Image Classification
Local Interpretable Model-Agnostic Explanations (LIME)
Grad-Cam
Proposed Methodology
Performance Measures in XAI
14.8 Experimental Result and Analysis
Datasets
14.8.1 Interpretability by LIME
Results of GradCam
14.9 Conclusion
References
Chapter 15
XAI in Health Care for Making Intelligent Decisions for COVID Prevention and Detection
Abstract
15.1 Introduction
15.2 Framework from Machine Learning XAI
15.2.1 Linear Regression
Applications
Advantages
Disadvantages
15.2.2 Random Forest
Uses of Random Forest
Applications of Random Forest
Steps for the Random Forest Algorithm:
Advantages
Discussion on Random Forest
15.3 COVID Data Set Description
15.3.1 Data Pre-Processing
The Major Steps for Data Pre-Processing
Knowledge Representation of XAI
Interpretability and Explainability
Explainability
15.4 XAI Approaches and Analysis in Sustainable Smart Health Care
Informatics
Characteristics of Explainable AI in Healthcare
15.5 XAI in Threats
Challenges in XAI
15.6 Experimental Analysis
15.7 Opportunities in XAI
15.8 Conclusion
References
Section 4. Medical Imaging Classification using Explainable AI
Chapter 16
Explainable AI in Healthcare
Abstract
16.1 Introduction
16.1.1 Interpretable AI
16.1.2 Transparent AI
16.1.3 Interactable AI
Objective of Study
16.2 General Process of XAI
16.3 Literature Survey
16.4 Proposed Approach with COVID-19 and Pneumonia Case Study
16.5 Problems Associated with Current XAI Techniques
16.6 XAI Recommandations System
16.7 Conclusion
References
Chapter 17
Medical Image-Based Steganography Using Deoxyribonucleic Acid (DNA) Algorithm
Abstract
17.1 Introduction
17.2 Literature Review
17.3 Proposed method
Embedding Procedure
17.4 The Process of this Proposed Model Takes Place in the Following Sequence of Steps
STEP 1 - DNA Encoding
Step 2 - Algorithmic Selector
Step 3 - LSB Substitution Method
Decoding Procedure
17.5 Experimental Results Analysis
17.6 Stego-Images Quality Analysis
Robustness Analysis
17.7 Comparative Analysis
17.8 Conclusion
References
Chapter 18
Deep Convolutional Neural Network for Classifying COVID-19 and Pneumonia Using Chest X-Ray Images
Abstract
18.1 Introduction
18.2 Related Works
18.3 Materials and Methods
18.3.1 Dataset Description
18.3.2 Convolutional Neural Network (CNN)
Convolution Layer
Pooling Layer
Fully Connected Layer
Output Layer
Pre-Trained Models
ResNet50 and ResNet101
VGG16 and VGG19
18.4 Performance Matrix for Classification
18.5 Results and Discussion
18.6 Conclusion
References
Chapter 19
XAI-Driven Visualization for Improved Decision-Making in Sustainable and Smart Healthcare
Abstract
19.1 Introduction
19.2 XAI-Driven Visualisation Techniques for Healthcare
19.2.1 Line and Area Charts
19.2.2 Bar Charts and Histogram
19.2.3 Scatter Plots
19.2.4 Heatmaps
19.3 Geographic Information System Mapping
19.3.1 Interactive Dashboard
19.3.2 Decision Tree
19.4 Applications of XAI-Driven Visualization in Sustainable Healthcare
19.4.1 Visualization for Sustainable Supply Chain Management
19.4.2 Visual Analytics for Sustainable Healthcare Planning
19.4.3 Applications of XAI-Driven Visualization in Smart Healthcare
19.4.4 Real-Time Monitoring and Visualization of Vital Signs
19.4.5 Visual Analytics for Population Health Management
19.5 Challenges and Considerations in Implementing XAI-Driven Visualization
19.5.1 Data Quality and Availability
19.5.2 Design and Usability
19.5.3 Privacy and Security
19.5.4 Ethical Considerations
19.6 Discussion
19.7 Conclusion and Future Work
References
Chapter 20
Threats, Difficulties and Possibilities for XAI in Healthcare Decision Support Systems
Abstract
20.1. Introduction
20.2. Fundamental Concepts and Background
20.2.1. Healthcare Decision Support Systems
20.2.2. Explainable AI (XAI)
20.2.2.1. The Need for XAI : Fair and Ethical Decision-Making
20.2.3. XAI in Medicine
20.2.4. Techniques and Explanation
20.2.4.1. Ante-Hoc Methods
20.3. Materials and Methods
20.3.1. Research Questions
20.4. Results
20.4.1. What AI-Based HDSS Have Been Developed That Incorporates XAI?
20.4.2. What Benefits Have Been Reported When Addressing Different Aspects of the Use of XAI in HDSS?
20.4.3. Is HDSS Expressed in Literature?
20.5. Discussion
20.5.1. Guidelines for Implementing Explainable Models in HDSS: Difficulties, Possibilities, and Future Research Needs
20.6. Conclusion
References
About the Editors
Index
Blank Page