Artificial Intelligence and the Fourth Industrial Revolution

Utpal Chakraborty, Amit Banerjee, Jayanta Kumar Saha, Niloy Sarkar, Chinmay Chakraborty, Utpal Chakraborty, Amit Banerjee, Jayanta Kumar Saha, Niloy Sarkar, Chinmay Chakraborty

• Publisher: Jenny Stanford Publishing
• Year: 2022
• Print ISBN:
  9789814800792
• eBook ISBN:
  9781000367935
• Topic: Computer Science
• Subtopic: Computer Science General
• Index: Computer Science

This book presents the overall technology spectrum in artificial intelligence (AI) and the Fourth Industrial Revolution, which is set to revolutionize the world. It discusses their various aspects and related case studies from industry, academics, administration, law, finance, and accounting as well as educational technology. The contributors, who are experts in their respective fields and from industry and academia, focus on a gesture-recognition prototype for specially abled people; jurisprudential approach to AI and legal reasoning; automated chatbot for autism spectrum disorder using AI assistance; Big Data analytics and Internet of Things (IoT); role of AI in advancement of drug discovery; development, opportunities, and challenges of the Fourth Industrial Revolution; legal, ethical, and policy implications of AI; Internet of Health Things for smart healthcare and digital wellbeing; machine learning and computer vision; computer vision-based system for automation and industrial applications; AI-IoT in home-based healthcare; and AI in super-precision human brain and spine surgery. Buttressed with comprehensive theoretical, methodological, well-established, and validated empirical examples, the book covers the interests of a broad audience from basic science to engineering and technology experts and learners. It will be greatly helpful for CEOs, entrepreneurs, academic leaders, researchers, and students of engineering, biomedicine, and master's programs in science as well as the vast workforce and students with technical or non-technical backgrounds. It also serves common public interest by presenting new methods to improve the quality of life in general, with a better integration into society.

Section II Internet of Medical Things (IoMT)

Chapter 4 Internet of Health Things: Opportunities and Challenges

DOI: 10.1201/9781003159742-6

Emeka Chukwu, Lalit Garg, and Ryan Zahra

Department of Computer Information System, University of Malta, Msida, Malta [email protected]

This chapter highlights the information revolution and the related opportunities and challenges. It specifically looks at the inherent problems with healthcare systems that stem from the presence of multiple actors and how digital systems have attempted to solve them. The critical problem of the shortage of skilled personnel at the primary healthcare level has led to undercollection or zero collection of vital statistics and biostatistics of patients. The chapter further discusses the emergence of autonomous systems with control systems and the full spectrum of digitization, with specific focus on the Internet of Health Things (IoHT). The key applications of IoHT are discussed, along with the key opportunities and pitfalls and what to look out for when architecting one. The architecture, network, power, and other design considerations of a maternal health use case model are discussed. The full prototype components of a maternal health self-service kiosk are described to illustrate this.

4.1 Introduction: Health System

Historically, across the globe, health has been about the patient getting well and remaining well. Most of the care provisions happen at the health facility, though trends are changing with the emergence of patient-centered care models [1]. There are over a hundred specialties and subspecialties in medicine and many more allied medical and support specialists, like nurses, midwives, pharmacists, and laboratory scientists and their subspecialists. This means that modern healthcare depends heavily on teamwork and communication and information [2]. Also given that healthcare has to do with human life, it is one of the heavily regulated sectors, just as it is complicated. The complex nature of this endeavor has made efficient, equitable, and effective healthcare delivery a “wicked” problem to solve for many countries. Specifically, many countries in the sub-Saharan African region still struggle with issues like preventable deaths of women in labor or immediately after delivery [3]. Routine health information in these countries is still aggregate-based, and other information is augmented by statistical analysis of extensive surveys.

4.1.1 Health Information Revolution

Traditional data collection, use, storage, and retrieval systems are paper based, though efforts are on to digitize these processes and records as much as possible. One significant limitation of the paper-based system is that only one person in the care continuum, in most cases at a different geographic location, can use the paper form or folder at any given time. It can even result in further delays if the patient has multiple reasons for visiting a care facility, as is often the case for the elderly. This limitation and other limitations of paper-based health and medical information have been identified with advances in technology.

Digitization of health facility information, specifically patient information, has led to an enormous amount of information from many different sources. Information is also provided by nontraditional sources, such as imagery equipment, genomic data, and environmental and body sensors. It is now commonplace to have patients access and manage information related to their own health through mobile applications.

4.1.2 Health Workforce and Task Shifting

The issues around human resources for health and their number, quality, and how to track and manage them have occupied the interests of policy makers across many jurisdictions. Recent findings from a few countries show that Africa still lags in terms of the required number and quality of health workers per citizen. Often, health facilities are staffed by untrained volunteers who perform more functions than they are supposed to because of the paucity of the workforce to do these tasks. This development even prompted some countries to recognize and try to formalize these task shift endeavors. Volunteers were provided support to do the work of nurses where there is a shortage of nurses, and midwives could perform some surgical procedures with little support and training. These efforts received a mixed response in terms of success and acceptance amid the cry for ethical consideration. This chapter advocates shifting tasks that machines can do to machines and that patients can do to them. This, it is anticipated, will free up more skilled professionals to deliver better care and help improve health outcomes where they matter most. Also, the data currently collected by these professionals often lack in quality and usability. In many cases, data are collected only for reporting, and no thought is given to closing the feedback loop in an actionable and patient-specific way; feedback happens at the aggregate level only. Audits are few and far between, but when they do happen, they focus on validating that the information entered into the summary form tallies with that in the patient registers. The registers are many, often one for each health program area. For example, there will be a register for HIV/AIDS, another for antenatal, and yet another for family planning.

4.1.3 Digitization: Hope, Hype, and Harm

Technology advances have given rise to a long list of promises that digitization will deliver, and these have been slow to arrive. Health systems in many parts of the world (especially the third world) remain unreliable, unsafe, unsatisfying, too expensive, and out of reach of many [4]. People have been waiting for when and where these promises will be delivered. Some have even taken the initiative to incentivize mass adoption, like the US government’s meaningful use project [5]. Many have developed national strategies, yet this often results in little progress. When this information should be collected, in what format, by whom, and where are some of the questions that currently limit the impact of the expected information revolution. Advances in technology have also been known to cause harm and sometimes death in the case of diagnostic systems. The privacy of patient information is attracting interest and sometimes poses impediments to digitization.

4.2 Internet of Health Things

The technology that the Internet of Health Things (IoHT) is based on is the Internet of Things (IoT), a technology that involves connecting several smart devices [6, 7]. The enabling technologies for IoT are mobile communication and wireless networks, which make up the core foundation. Near-field communication (NFC) and radio-frequency identification (RFID) are two of the fundamental technologies that enable IoT [8]. The NFC technology came to fruition as the RFID technology evolved. Nowadays, it is also becoming a norm to include the NFC technology in smartphones, and this offers a different type of wireless connectivity than that offered by the more common Bluetooth [6]. These technologies make it easier and more efficient for devices to collect more substantial volumes of data than was previously possible from different devices and analyze it [9]. The same concept has also been applied to the area of healthcare in many different ways but always to allow smart objects to communicate and therefore create an ecosystem that can provide better healthcare [10].

Before moving forward in this chapter, it is vital to discuss the history of IoT to clarify IoHT, a new term not as popular as IoT but whose popularity is already growing. One may have heard of terms like Web of Things, WoT in short, and Internet of Medical Things, also known as IoMT. This chapter is focused on IoHT. Borrowing the definition from Guinard and Trifa, the “Internet of Things is a system of physical objects that can be discovered, monitored, controlled, or interacted with by electronic devices that communicate over various networking interfaces and eventually can be connected to the wider Internet” [11]. One will agree that many of things that featured in the definition are already around us; this was not the case two decades ago.

Our everyday lives can be extended through the use of digital augmentations. The actual breakthrough was mentioned in the 1965 paper by Intel’s chairman emeritus, Gordon Earle Moore, where he predicted that the computing power (or, more precisely, the number of transistors on a silicon wafer) for a given size would double every two years [12]. This prediction seems to have held for the last five years, resulting in microprocessor-controlled devices unprecedentedly small in size. This miniaturization has led to many advances and has increased the world of what is now possible. Today, we live in a world of wearables and embedded sensing devises. Around us, washing machines, televisions, fridges, air conditioners, and anything one can think of can be digitally controlled. When these things are connected, we have the Internet (or Intranet), which completes the IoT paradigm. At this point, it is essential to distinguish between the Internet and an intranet, though they both contribute to complete the IoT paradigm. An intranet is a functional network that does not necessarily connect to the Internet. The Internet, on the other hand, is a network that somehow allows access to the Internet worldwide Web.

Describing the different protocols and components that make up the Internet part is well beyond the scope of this one chapter. One clear thing is that our Bluetooth, Wi-Fi, ZigBee, 2G, 3G, 4G, 5G, infrared, and cable connected protocols all make up the network suite. A Google trend analysis shows that though the term was coined sometime in 2009, it did not generate global interest until after 2013, trigged likely by Google’s purchase of NEST (an IoT company) for US$3.3 billion in December 2013. The growing interest is shown by the Google trend analysis in Fig. 4.1.

Figure 4.1 Google trend analysis: Internet of Things.

4.2.1 Opportunities

The use cases of IoT in healthcare are growing, with new opportunities for application emerging. While the existence of advanced digital solutions in a health system may make adoption easy, it can often be a barrier to entry. Sometimes, not having existing digital systems, as is the case in most African health systems, can be an opportunity as there is nothing to tear down because nothing currently exists. The opportunities continue to grow with autonomous systems in which machines are now increasingly talking to each other. The IoT market will be worth about US$1.7 trillion in 2020, with Cisco suggesting that the number of connected devices will grow to 50 billion around the same time [13]. This scenario will be in a world with less than 8 billion people, which means that in a few months, there will be four connected devices per individual. This number may even be higher if the number of communities not currently connected is considered.

IoHT brings about several advantages, one of which is that the burden on the nurses to collect and analyze the vast amounts of data is taken care of by this technology. Before such technologies were adopted, the nurses had to collect the patient data...

Table of contents

• Cover Page
• Half Title Page
• Title Page
• Copyright Page
• Table of Contents
• Preface
• Section I AI in Industry 4.0
• Section II Internet of Medical Things (IoMT)
• Index