BookNiklas Lidströmer, Yonina C. Eldar, editors.
Summary: This book deals with the advantages of using artificial intelligence (AI) in the fight against the COVID-19 and against future pandemics that could threat humanity and our environment. This book is a practical, scientific and clinically relevant example of how medicine and mathematics will fuse in the 2020s, out of external pandemic pressure and out of scientific evolutionary necessity. This book contains a unique blend of the world's leading researchers, both in medicine, mathematics, computer science, clinical and preclinical medicine, and presents the research front of the usage of AI against pandemics. Equipped with this book the reader will learn about the latest AI advances against COVID-19, and how mathematics and algorithms can aid in preventing its spreading course, treatments, diagnostics, vaccines, clinical management and future evolution.
Contents:
Intro
Foreword
Preface
Contents
About the Editors
Chapter 1: Introduction to Artificial Intelligence in COVID-19
Pandemics
History of Pandemics
The COVID-19 Pandemic
Origins of the COVID-19 Pandemic
Continuous Fight for Science and Reason
Modern Tools for Pandemic Control
A Brief Chronology of the Chapters of This Book
Power of Science
References
Chapter 2: AI for Pooled Testing of COVID-19 Samples
Introduction
System Model
The PCR Process
Mathematical Model
Pooled COVID-19 Tests
Recovery from Pooled Tests Group Testing Methods for COVID-19
Adaptive GT Methods
Non-Adaptive GT Methods
Pooling Matrix
Noiseless Linear Non-Adaptive Recovery
Noisy Non-Linear Non-Adaptive Recovery
Summary
Compressed Sensing for Pooled Testing for COVID-19
Compressed Sensing Forward Model for Pooled RT-PCR
CS Algorithms for Recovery
Details of Algorithms
Assessment of Algorithm Performance and Experimental Protocols
Choice of Pooling Matrices
Choice of Number of Pools
Use of Side Information in Pooled Inference
Comparative Discussion and Summary
References Chapter 3: AI for Drug Repurposing in the Pandemic Response
Introduction
Desirable Features of AI for Drug Repurposing in Pandemic Response
Technical Flexibility and Efficiency
Clinical Applicability and Acceptability
Major AI Applications for Drug Repurposing in Response to COVID-19
Knowledge Mining
Network-Based Analysis
In Silico Modelling
IDentif.AI Platform for Rapid Identification of Drug Combinations
Project IDentif.AI
IDentif.AI for Drug Optimization Against SARS-CoV-2
IDentif.AI 2.0 Platform in an Evolving Pandemic IDentif.AI as a Pandemic Preparedness Platform
Use of Real-World Data to Identify Potential Targets for Drug Repurposing
Future Directions
References
Chapter 4: AI and Point of Care Image Analysis for COVID-19
Introduction
Motivation for Using Imaging
Motivation for Using AI with Imaging
Integration of Imaging with Other Modalities
Literature Overview
Chest X-Ray Imaging
Diagnosis Models
Prognosis Models
Use of Longitudinal Imaging
Fusion with Other Data Modalities
Common Issues with AI and Chest X-Ray Imaging
Duplication and Quality Issues Source Issues
Frankenstein Datasets
Implicit Biases in the Source Data
Artificial Limitations Due to Transfer Learning
Computed Tomography Imaging
Diagnosis Models
Prognosis Models
Applications to Regions Away from the Lungs
Use of Longitudinal Imaging
Fusion with Other Data Modalities
Common Issues with AI and Computed Tomography Imaging
Ultrasound Imaging
What Can be Observed in LUS
Models Assisting in Interpreting LUS
Diagnosis Models
Prognosis Models
Use of Longitudinal Imaging
Common Issues with AI and Ultrasound Imaging
Conclusions