Front Cover
Artificial Intelligence in Healthcare and COVID-19
Copyright Page
Contents
List of contributors
Preface
1 Improvement of mental health of frontline healthcare workers during COVID-19 pandemic using artificial intelligence
Other notes
1.1 Introduction
1.2 Background
1.3 Main content
1.4 Methodologies and implementation
1.5 Discussion
1.5.1 Connection to artificial intelligence
1.5.2 Strengths
1.5.3 Weaknesses
1.6 Conclusion
References
2 Effective algorithms for solving statistical problems posed by COVID-19 pandemic
2.1 Introduction
2.2 Forecasting the epidemic curves of coronavirus
2.2.1 Forecasting models for the COVID-19 outbreak
2.3 Nonparametric tests used for forecasting models estimation
2.3.1 Nonparametric tests for homogeneity
2.3.2 Exact nonparametric test for homogeneity
2.4 Comparison of forecast models
2.5 Conclusion and scope for the future work
References
3 Reconsideration of drug repurposing through artificial intelligence program for the treatment of the novel coronavirus
3.1 Introduction
3.2 Viral morphology
3.2.1 Structured proteins
3.2.1.1 Spike protein/spike membrane
3.2.1.2 Membranous proteins
3.2.1.3 Nucleic acid-protein/nucleocapsid
3.2.1.4 Enveloped protein
3.2.2 Nonstructured proteins
3.2.2.1 Proteases
3.2.2.2 RNA-dependent polymerase
3.2.2.3 Helicase
3.3 Virus lifecycle
3.3.1 Life process of severe acute respiratory syndrome 2
3.3.1.1 Attachment and entry
3.3.1.2 Replication and transcription
3.3.1.3 Assembly and release
3.4 Currently available viral targeting drug candidates at various stages of life cycle
3.5 Different drug repurposing approaches
3.5.1 Target approach
3.5.2 Knowledge-dependent approach
3.5.3 Molecular docking-based approach.
3.5.4 Machine learning approaches
3.5.5 Pathway-based approaches
3.5.6 Artificial neuronal network approaches
3.5.7 Deep learning machine approaches
3.5.8 Network modeling approach
3.5.8.1 Autoencoder approaches
3.5.8.2 Text mining approaches
3.6 Artificial intelligence algorithms for drug repurposing
3.7 Computational intelligence-based approaches to identify therapeutic candidates for repurposing against coronavirus
3.7.1 Network-based model
3.7.2 Structure-based approaches
3.7.3 Artificial intelligence approaches
3.8 Challenges in drug repurposing
3.9 Future perspectives of artificial intelligence-informed drug repurposing
3.10 Conclusion
References
4 COVID-19: artificial intelligence solutions, prediction with country cluster analysis, and time-series forecasting
4.1 Introduction
4.1.1 Motivation for this study
4.1.2 Adverse impacts of COVID-19 outbreak
4.1.3 Chapter organization
4.1.4 Table of acronyms used in this chapter
4.2 Review of literature on COVID-19 pandemic
4.3 K-means clustering for COVID-19 country analysis
4.3.1 Cluster analysis: an overview
4.3.2 Dataset selection and preprocessing
4.3.3 Findings from COVID-19 country cluster data analysis
4.3.4 The results and discussions
4.4 Time-series modeling for COVID-19 new cases
4.4.1 Time-series modeling: an overview
4.4.2 Dataset description
4.4.3 Time-series exploration
4.4.4 Predictive analytics
4.5 Conclusion
References
Further reading
5 Graph convolutional networks for pain detection via telehealth
5.1 Introduction
5.2 Methodology
5.2.1 Features extraction
5.2.2 Graph-based modules
5.2.3 Frame-wise weight calculation
5.2.4 Classification
5.3 Experiments
5.3.1 Datasets
5.3.2 Experimental setting
5.4 Results and discussion
5.5 Conclusion.
Acknowledgment
References
6 The role of social media in the battle against COVID-19
6.1 Introduction
6.2 Materials and methods
6.3 Related reviews
6.4 Understanding COVID-19 data
6.4.1 Topic detection
6.4.2 Sentiment analysis
6.5 Misinformation identification and spreading
6.6 COVID-19 forecasting
6.7 Discussion: challenges and future directions
6.8 Conclusion
References
7 De-identification techniques to preserve privacy in medical records
7.1 Introduction
7.2 Background
7.2.1 Deep learning systems
7.2.2 Language models and embeddings
7.2.3 Clinical de-identification, low-resource languages, and transfer learning
7.3 Material and methods
7.3.1 Data sets
7.3.1.1 The SIRM COVID-19 de-identification corpus
7.3.1.2 The i2b2/UTHealth 2014 de-identification corpus
7.3.2 System architectures
7.3.2.1 BiLSTM plus CRF-based architecture
7.3.2.1.1 Embedding layer
7.3.2.2 BERT-based architecture
7.3.3 Experimental setups
7.3.3.1 BiLSTM plus CRF-based systems
7.3.3.2 BERT-based systems
7.3.4 Evaluation metrics
7.3.5 Training strategies
7.4 Results and discussion
7.5 Conclusion
References
8 Estimation of COVID-19 fatality associated with different SARS-CoV-2 variants
8.1 Introduction
8.1.1 Related work
8.2 Materials and methods
8.2.1 Data on COVID-19 infections and deaths
8.2.2 Data about SARS-CoV-2 variants
8.2.3 Models to estimate fatality
8.2.4 Uncertainty of available data and fatality estimation
8.2.5 Correlation with vaccine distribution
8.2.6 Hypotheses to generalize conclusions
8.3 Results
8.4 Discussion and conclusion
References
9 Artificial intelligence for chest imaging against COVID-19: an insight into image segmentation methods
9.1 Introduction
9.2 Chest CT findings of COVID-19 pneumonia.
9.3 Medical image segmentation and artificial intelligence
9.3.1 The fourth generation of segmentation methods: deep learning approaches
9.3.2 Evaluation metrics
9.4 Existing methods for COVID-19 chest CT images segmentation
9.4.1 Lung-region-oriented methods
9.4.2 Lung-lesion-oriented methods
9.4.2.1 Binary lung lesion methods
9.4.2.2 Multi-class lung lesion methods
9.5 Attention-FCNN: a novel DL model for the segmentation of COVID-19 chest CT scans
9.5.1 Chest CT imaging dataset
9.5.2 Attention-FCNN architecture
9.5.3 Attention gates: structure and functioning
9.5.4 Training details
9.5.5 Results
9.5.6 Ablation study
9.6 Discussion and conclusions
References
Index
Back Cover.