Computing Reviews

Computational intelligence techniquess for combating Covid-19
Kautish S., Peng S., Obaid A., Springer International Publishing,Cham, Switzerland,2021. 593 pp.Type:Book
Date Reviewed: 01/16/23

This book looks at the application of computing techniques, especially intelligence techniques. Machine learning (ML) techniques have been developed for many decades now and the practical applications of ML have significantly increased over time. The Covid-19 virus has directly (or indirectly) affected almost every part of the world. The book has 17 research chapters on ML and the application of relevant techniques for Covid-19. To combat Covid-19, ML is used for diagnosis, prevention, and risk assessment. ML techniques such as clustering, classification, deep learning, and so on are applied.

The book cannot be used as a textbook. It is a good source for artificial intelligence (AI) or healthcare students looking for graduate-level project ideas. It is also good reading for researchers in AI applications and healthcare.

Readers with a general interest can check out the survey papers. Chapter 3’s clinical perspective uses a block diagram to compare AI-based techniques versus traditional tests for detecting Covid-19 patients. It also includes a table of major AI software being used to combat Covid-19. Then chapter 5 provides a survey of various Covid-19 methods from a healthcare perspective; each method’s advantages and disadvantages are discussed. Chapter 7’s survey on the Internet of Things (IoT) shows devices that can be used for healthcare and telemedicine. The proposed Internet of Covid-19 things (IoCT) includes AI and blockchain applications. Chapter 15, also on IoT applications, describes a random pooling algorithm for data analysis. Chapter 14 describes chatbot-related applications for combating Covid-19, and chapter 9 is devoted to vaccine finding procedures. Another chapter from a purely healthcare perspective describes the immunization approach used.

Chapter 1 carries out a risk analysis on Covid-19 cases. Available datasets are death rates by country and risk factors by country. The weights of risk factors leading to death are found. Using the weight matrix, countries are clustered into high risk, low risk, and medium risk. The conclusion is that South Asian countries have low death rates because of low risk factors. The paper presents its methodology and the multiple analytic hierarchy process as core techniques.

Chapter 4 presents another type of risk analysis. The risk factors are symptoms such as body temperature, drowsiness, cough, weakness, and so on. The C4.5 ML algorithm (an algorithm based on decision algorithms) is applied. The paper reports a 75 percent accuracy for the prediction of cases. Chapter 10 shows how pattern recognition can be used for personalized treatments. Covid-19 patterns can be detected using analytics technologies on big data. The chapter refers to molecular dynamics simulation and immunization dynamics simulation. Chapter 13 also looks at factors affecting Covid-19 cases; here data from four cities is collected and ML techniques are applied to predict infection rates.

Various fields apply deep learning using neural networks for image processing applications. Chapter 2 uses convolutional neural networks (CNNs) to classify Covid-19 cases using computerized tomography (CT) scans and X-ray images. X-ray machines have high availability in countries struggling to find cheaper ways to identify Covid-19 cases. CNNs can learn from existing Covid-19 cases and then classify patients as normal patients, pneumonia patients, or Covid-19 patients. Chapters 12 and 17 provide more details about using deep learning on X-Ray images. Chapter 6 uses CNNs for detecting social distancing--social distancing is recognized as an effective method for preventing the fast spread of Covid-19. The code for the application detects objects and then finds if the distance between them is smaller than the required distance. This code is implemented on Jetson Nano, a computer that can run CNNs in parallel for image processing. Chapter 11 uses deep learning in the treatment of depression and anxiety. It shows how augmented reality/virtual reality (AR/VR) can be used.

On the socio-economic front, the optimal distribution of techniques for combating Covid-19 is very important. Chapter 8 presents an optimizing algorithm for such distribution. Chapter 18 analyzes the effects of information sharing on various social and economic fronts. Surprisingly, it talks about the negative impacts of information sharing, for example, too much misinformation.

Reviewer:  Maulik A. Dave Review #: CR147537

Reproduction in whole or in part without permission is prohibited.   Copyright 2023 ComputingReviews.com™
Terms of Use
| Privacy Policy