Abstract
Due to the increase in healthcare data provided by IoT, there is a need to use new methods for data analysis. Machine learning (ML) techniques promise solutions for many challenges facing the IoT-based healthcare services. MLs provide significant improvement in different IoT aspects related to storage size, computational power, and data transfer speeds. In addition, MLs provide a number of solutions for medical imaging, resources, medical data processing, detection, diagnosis, and prediction. Recently, many applications have appeared in the field of medicine and healthcare, which are closely related to the IoT. This chapter presents basic concepts related to the use of ML techniques, in addition to some algorithms used in the medical field and healthcare technology based on IoT devices and systems. Moreover, the chapter will discuss the ML opportunities and challenges in healthcare and future trends as well. The chapter gives the reader full perception of the possibility of using ML techniques in the medical and healthcare fields, with a systematic description of their applications.
TopIntroduction
IoT aims to allow objects/things to connect with each other at any place and time, with any network to do any service for anyone (Alnazir et al. 2021) and (elmustafa el at. 2021). IoT can be defined as a group of connected manageable sensors or actuators representing a part of infrastructure with the ability to connect with the other network devices and exchange collected data to achieve a specific function (Bruno et al. 2015). With the development of the IoT and entry in many areas, its use area has expanded to include many applications. IoT applications include; supply chain management, transportation, disaster alerting & recovery, and healthcare applications.
Nowadays, healthcare represents the most important aspect of people’s lives, so most countries strive to provide a strong and effective health care system. Technologies such as AI, and IoT, have entered the healthcare field, turning traditional healthcare into intelligent healthcare (Zeinab et al. 2017). IoT-based healthcare systems provide a promising solution for the inefficiencies in health infrastructure by generating a global network of smart connected devices (Stephanie et al. 2017). These devices enable the remote monitoring of the patients and continuously process the data and store it in the cloud to be ready when it is needed in additional processes
The world is witnessing an increase in the number of births on the one hand and the aging populations on the other hand, which has led to an increase in the number of patients with chronic diseases (Australian et al. 2014). This increasing causing a huge pressure on healthcare system. Moreover, population healthcare occupies large amount of the countries budgets. Effective solutions must be obtained to reduce pressure on health facilities and staff as well as increasing the quality of provided health services (Stephanie et al, 2017). Due to being, an important solution IoT interred into a number of healthcare applications domains. The IoT based Healthcare Network as shown in figure 1, enables healthcare devices to communicate with the IoT through gateways in the communication network and to send and exchange the healthcare information to the health smart decisions and cloud and data platforms for analytics and applications services. In general, IoT healthcare applications can be categorized into three scenarios as reviewed in the next subsections.
Figure 1. IoT based healthcare network
Key Terms in this Chapter
Acute Diseases: Know the diseases that evolve suddenly and at a short time such as fast infection, accident injuries, and poor drugs. Such diseases are attached specific symptoms, require urgent health care, and are improved as soon as the urgent intervention of appropriate treatment.
E-Health: The concept is related to the health care that supported by electronics and communications technology, where it reflects the use of Internet, medical hardware, and computer relationship with the request in addition to health informatics. The term also refers to health services, information provided or strengthened through the Internet and related techniques.
Representational State Transfer (REST): An architectural way of building software that is designed to develop a IoT Web Architecture for Monitoring Field that defines a number of constraints on how the Internet's distributed hypermedia system architecture behaves. It also enables scalability in interactions between components, unified interfaces, and the creation of a multi-layered architecture for storing components.
Internet of Medical Things (IoMT): A technology that connects a many medical devices and applications with information technology systems online related to healthcare. It can also be described as the possibility of allowing medical devices that contain communication mechanisms such Wi-Fi to connect with other devices to the Internet and enable the storage of medical data through different cloud platforms.
Chronic Diseases: Such diseases appear after a while, where they begin to slowly evolve and long term. These diseases appear from non-healthy lagging behaviors such as malignancy, physical idle, or overlooking specific materials for a long time. Some are classified as addictive symptoms such as smoking and others because of the age of patient, such as Alzheimer, diabetes, and heart disease. Such diseases need a special health pattern and exercise a specific healthy behavior with continuous monitoring.
Remote Health Monitoring (RHM): This term means providing health information related to patients. The techniques used to supply different data to the patient's equipment managers, and provide location and data analysis capabilities. This concept is also possible to deliver data through cellular communications to health centers or doctors, regardless of where it is for medical equipment.
Service-Oriented Architecture (SOA): An architectural method that supports the routing of services provided through a network communication protocol. SOA enables to pass messages using description metadata through protocols services. It is also enables exchanging the information between similar and dissimilar applications and solve the problem of interoperability by using an external web services.