Chapter 1
Internet of Things, Edge, Fog, and Data Analytics in Smart Healthcare: Introduction, Benefits, and Challenges

After reading this chapter, you should be able to:

• Understand the use of Internet of Things, edge, and fog computing in smart healthcare.
• Understand components and applications of big data analytics for healthcare.
• Understand how to apply various data analysis schemes for some use cases (i) forecasting tuberculosis (TB) prevalence rates, (ii) skin aging estimation.
• Identify current challenges and future directions in smart healthcare.

1.1 Introduction

Throughout history, advancements in science and technology have often been driven by the need for medical and healthcare applications. The emergence of cloud technology has revolutionized healthcare information systems by providing infrastructure, platforms, and software as a service. For decades, cloud technology has played a dominant role in these systems. However, one significant limitation of cloud-based applications is their high service response time. In an emergency situations, where rapid decision-making and real-time monitoring of a patient's condition are critical, delays can severely impact patient outcomes. To address these challenges, innovative computing solutions such as cloud computing, edge computing, and fog computing have been developed to optimize performance and reduce response times in healthcare scenarios. In this chapter, we compare various computing technologies and introduce a unified architectural framework for Internet of Health Things (IoHT) applications based on fog computing. We also explore potential applications and challenges associated with integrating fog computing into Internet of Things (IoT) healthcare systems. This chapter also highlights the significant potential of fog computing for enhancing IoT-based health applications.

In the early 2000s, the introduction of fifth-generation mobile networks, commonly known as 5G, revolutionized various industries. 5G enables the delivery of services with ultra-low latency and ultra-high bandwidth, making it a game-changer for the IoT. This advancement has paved the way for a wide range of smart applications that benefit humanity, such as smart healthcare [1]. In a typical smart healthcare system, sensors are attached to a patient's body to monitor various health metrics, such as blood pressure, heart rate, and mobility. These devices continuously gather data and transmit it to a centralized system via network connections for further analysis and monitoring.

Due to the large volume of data collected, it is typically stored in the cloud for analysis and decision-making purposes [2]. However, a significant drawback of cloud-based applications is their high service response time [3]. In critical situations such as ambulance emergencies, where rapid data processing is essential - encompassing factors like hospital location, distance, available ambulance vehicles, and medical staff - decisions must be made in a matter of seconds, as delays can have life-threatening consequences. In such scenarios, cloud computing may not provide the necessary real-time responsiveness. To address this issue, fog computing and edge computing have been proposed as solutions [4, 5]. Fog computing extends cloud capabilities to the local network, thereby significantly reducing service response time and improving the ability to make real-time decisions.

Globally, over 200 million people require regular monitoring due to chronic conditions such as cancer, asthma, cardiovascular disease, arthritis, dementia, Alzheimer's, visual impairment, and chronic obstructive pulmonary disease [6, 7]. This number is growing daily, necessitating the use of various technologies for processing. IoT devices, integrated with sensors in healthcare systems, facilitate automated patient monitoring, activity tracking, heart rate detection, and caloric expenditure/intake calculation, among other functions. The data produced by these IoT devices are processed and analyzed either at fog/edge devices or cloud data centers. However, current cloud models may not be the most effective for addressing IoT challenges due to their high transmission requirements, dependency on structured frameworks, and slow response times, which make them less suitable for critical applications. Another challenge is determining what to offload - data, computation, or applications - and where to offload it-fog or cloud-and to what extent. Given task-related factors such as size, duration, arrival rate, and resource needs, fog-cloud collaboration can be unpredictable. Effective dynamic task offloading is essential to optimize the use of fog and cloud resources [8]. Fog computing in conjunction with IoT addresses these needs [9], leading to significant improvements in the healthcare system by reducing data volume and network bandwidth requirements [10].

1.2 Use of Edge and Fog Computing for Healthcare Applications

Fog computing, a component of cloud computing situated closer to the end user, enhances user efficiency, authenticity, and usability. It provides a space for data storage, computation, and communication with edge devices, thereby improving real-time privacy and security [11]. Fog computing operates between the cloud and user devices [12], and it is increasingly applied across various domains, including smart homes, industries, and hospitals. In the context of smart hospitals, many researchers have proposed and developed different architectures for implementing fog computing [13, 14]. Numerous studies have reviewed these proposals and designed various architectures to illustrate the fundamental principles of fog computing in healthcare [15, 16]. The cloud-fog healthcare architecture consists of three layers: (i) the IoT/sensor layer, (ii) the fog layer, and (iii) the cloud layer, as illustrated in Figure 1.1. The architecture typically processes fog data to reduce its volume, enhance flexibility and security, and subsequently transfer it to the cloud [17, 18]. The body sensor network monitors the patient's physiological parameters, including blood pressure, pulse rate, body temperature, pressure rate, electrocardiogram, and electroencephalogram. Wearable sensors continuously track these parameters and transmit the data to the fog layer via wireless networks like Bluetooth, Zigbee, IEEE 802.11, and WiMAX [19, 20]. The fog layer processes this data and sends health alerts to relevant parties, such as family members, caregivers, and authorized medical professionals, who can monitor vital signs through various applications [15, 21]. The patient's medical data is regularly aggregated and forwarded to cloud servers for further analysis. In the medical field, the integration of fog computing with IoT presents unique advantages for health monitoring systems.

Figure 1.1 Cloud-fog healthcare architecture.

Recently, the rapid growth of healthcare IoT applications leveraging fog computing has highlighted its vast potential. Hypertension is a serious condition that often affects the elderly or individuals with a history of the disease. It can cause sudden attacks, and without prompt intervention from family and medical professionals, the patient's life could be at risk. To address this issue, Sood et al. [22] proposed an IoT healthcare application utilizing fog computing to monitor and diagnose hypertension attacks in real-time. The system operates by using sensors attached to the patient to collect blood pressure data. This data is then analyzed by an artificial neural network, facilitated by fog computing, to predict potential blood pressure attacks. The system continuously sends blood pressure alerts to pre-configured mobile devices. Experimental results showed that this solution reduced service response time and bandwidth consumption while improving diagnostic accuracy compared to existing methods. In addition to this, Wang et al. [23] introduced a healthcare IoT system utilizing fog computing to minimize the time required for retrieving patient data. Their approach involves leveraging caching and request aggregation within content-centric networking. Experimental results demonstrated that their solution reduced patient data retrieval time by up to 28.5% compared to existing methods.

However, these applications also pose significant challenges in safeguarding patient health information's privacy and security. To address these issues, Saha et al. [24] introduced a privacy-focused e-healthcare framework for electronic medical records within fog computing-based healthcare IoT applications. Experimental results showed that their solution effectively improved privacy and security for patient information and reduced response times compared to existing approaches.

Recognizing the limitations of cloud-based healthcare IoT applications, Awaisi et al. [25] proposed a more efficient healthcare IoT framework based on fog computing to enhance both efficiency and security. Their proposed architecture includes an identity-based user authentication method to further bolster security. Experimental findings demonstrated that this solution improved performance and enhanced security.

The integration of fog computing into healthcare IoT applications is becoming an essential trend for enhancing real-time eHealth systems. However, the proliferation of IoT devices can lead to system overload. To address this issue, Zhang...