Читать книгу Smart Healthcare System Design - Группа авторов - Страница 12

The Internet-of-Things (IoT) interconnects humans with uniquely identifiable embedded computing devices within the existing internet infrastructure. It has emerged as a powerful and promising technology, and though it has significant technological, social, and economic impacts, it also poses new security and privacy challenges. Compared with the traditional internet, the IoT has various embedded devices, mobile devices, a server, and the cloud, with different capabilities to support multiple services. The pervasiveness of these devices represents a huge attack surface. And since the IoT connects cyberspace to physical space, known as a cyber-physical system, IoT attacks not only have an impact on information systems, but also affect physical infrastructure, the environment, and even human security. Nowadays, the IoT has received massive attention for applications in different domains, the healthcare sector being one of them. A healthcare system serves society by taking care of its citizens’ physiological and neurological conditions through sensors by amassing information on their current health conditions and passing it along to the healthcare center for necessary actions. Accordingly, physicians can examine these health conditions and take the steps required to prevent the deterioration of the patient’s health. The purpose of this book is to help achieve a better integration between the work of researchers and practitioners in a single medium for capturing state-of-the-art IoT solutions in healthcare applications to address how to improve the proficiency of wireless sensor networks (WSNs) in healthcare. It explores possible automated solutions in everyday life, including the structures of healthcare systems built to handle large amounts of data, thereby improving clinical decisions; which is why this book will prove invaluable to professionals who want to increase their understanding of recent challenges in the IoT-enabled health-care domain. The separate chapters herein address various aspects of the IoT system, such as design challenges, theory, various protocols, and implementation issues, and also include several case studies. Furthermore, this book has been designed for both undergraduate students and researchers to easily understand and apply IoT in the healthcare domain.

About the Book

Smart Healthcare System: Security and Privacy Aspects covers the introduction, development, and applications of smart healthcare models that represent the current state-of-the-art of various domains. The primary focus will be on theory, algorithms, and their implementation targeted at real-world problems. It will deal with different applications to give the practitioner a flavor of how IoT architectures are designed and introduced into various situations. More particularly, this volume consists of 14 chapters contributed by authors well-versed in the subject who are devoted to reporting the latest findings on smart healthcare system design.

Chapter 1 explores a framework that can use real-time electroencephalogram (EEG) signals from multiple channels to predict the occurrence of an epileptic seizure. A selected number of EEG channels are provided as input to the system, and the corresponding epileptic seizure state is recorded at every second. A hybrid artificial neural network with a support vector machine-based classification is created as a simulation of real-time dynamic predictions in this system.

Chapter 2 discusses the critical factors to be considered in mHealth applications, such as mobility awareness, location-based medication, data, distance, and measurement protection for eHealth. Most of the mHealth apps operate with the patient’s background, which involves disease and environmental observation. Many problems face creating these applications, such as protection, smartness, decision-making, application size, and timely actions. This study presents the health sector dilemma by using it fuzzy logic for changes in health. For the health application to enhance well-being, all features addressed in this chapter are imperative.

Chapter 3 includes the design of a decision-making framework that gathers, preprocesses, and analyzes data from IoT-based healthcare systems and produces comprehensive information reports for better diagnosis. It implements data preprocessing methods, such as data washing, munging, normalization, elimination, and noisy data removal. The integration of the IoT with data analytics technologies results in healthcare systems becoming smarter and smarter. In the preliminary stage alone, the collected IoT data, such as pulse rate, temperature, oxygen level, and heart rate from connected devices, can be used to analyze the need and severity using appropriate machine learning techniques. Multi-criteria decision-making (MCDM) strategies, such as SMART, WPM, and TOPSIS, are often used to create comprehensive, insightful diagnostic reports at the conclusion of the development process.

In Chapter 4, the proposed work deals with touch and native voice-assisted prototype design and development to allow intuitive communication and interaction between health professionals and patients affected by severe acute respiratory infection (SARI), who are dependent on a ventilator and admitted for quarantine treatment. It also ensures that the multilingual capacity to communicate effectively in most of the ten Indian languages is established so that patients are relieved of pain, etc., as health professionals answer their queries. Touch-based gesture patterns can be effectively used as an interactive module in this prototype and let doctors frequently track and react to ICU patient inquiries by updating it to easily communicate the patient’s emotions or pains to caregivers. The planned prototype would be made available and public in an open source software repository.

Chapter 5 discusses the critical importance, especially in developing countries, of identifying the cause of a pandemic, such as COVID-19, and monitoring the spread of the disease. Included in our proposed system presented in this chapter is a network model that incorporates wireless body sensors, wearable devices, and cloud computing to manage patient data in the form of text or images, or cloud voice. To keep track of the real-time data, a cell phone application is installed along with a website.

In Chapter 6, Healthcare 4.0 technologies are adopted so that patients can be tracked remotely for surgical operations. Biosensors are also adopted in handheld gadgets. The proposed framework uses machine learning techniques to analyze the data obtained by the sensors. This method gathers the medical records of patients for review. It is challenging to provide a bed for treatment in the current COVID-19 pandemic situation, especially in developing and highly populated countries. Thus, the proposed Healthcare 4.0 system is designed to move therapies with a high-precision disease detection rate and testing from hospitals to patients’ homes.

Chapter 7 explains why even though smart technology offers several healthcare benefits, the same systems have a more significant effect on both confidentiality and security. Hacks on other frameworks, personal security risks, privacy threats, data eavesdropping, etc., are potential threats. Therefore, together with a cloud server, the framework proposed in this chapter uses the wireless body area network (WBAN) to hold patients’ records and make them available to only the individuals concerned by creating a role-based assignment and least privilege access system. It gathers the medical history of patients for potential reference.

In Chapter 8, the proposed system is a fully automated diet monitoring solution consisting of food quality assessment sensors operated by Wi-Fi and a smart-phone application that collects nutrition information about food ingredients. The food weighing sensor calculates the food’s weight, which is transmitted to the cloud over the internet via a microcontroller integrated with wireless module synchronization that is included in the monitoring system. To achieve the required nutrient values, two separate approaches are used. The first process is an optical character recognition (OCR) process which tests the nutrient value using the FDA-mandated nutrition facts label. In the other process, the barcode of the food is scanned, and nutritional data is collected from the internet using free application programming interfaces (APIs). Food is thus categorized based on the highest nutritional value, the relationship between the food consumed, and the lack of nutrients.

Chapter 9 discusses the gradually increasing usage of smart devices in various domains, with a particular focus on fusing the IoT into the medical sector to enhance clinical consideration based on the patient. Maintaining the protection of the information generated and obtained by IoT devices is the most severe problem in administering medical services, so the main objective of this chapter is to establish a system for safeguarding the IoT data developed in medical services. Security mechanisms used in the IoT setting must also communicate from end to end and must be adopted by low-cost IoT devices.

Chapter 10 explores why the energy consumption of WSNs and IoT devices is considered to be the aggregation and transmission of data. In processing and transmitting redundant and unnecessary data, these devices waste their power. Therefore, this chapter presents a means of eliminating redundant data and reducing the number of data transmissions, thus reducing the energy consumption of the IoT devices. Also included is an end-user remote monitoring system that monitors and verifies performance during real-time communication of these smart objects.

Chapter 11 explores the stability, data storage, and performance of various IoT devices that reflect the disadvantages of integrating these kinds of tools in the business sector. Data on the cloud server is more often than not compromised, and data storage is inefficient due to the growing number of users and devices on the internet. However, as new ways of using the IoT are taking shape, these disadvantages must be rectified. The world awaits many developments in the coming decades that will gracefully upgrade current systems; for instance, the advent of edge computing will transubstantiate cloud computing by eliminating technicalities while retaining the appropriate use of bandwidth for data privacy. Besides which, the IoT is bound to change industries, healthcare, traffic control, cyber-security, etc. With its success and steady progress, the future of the IoT is auspicious, with the intent of paving a new path for technological growth. This chapter’s focus is on current IoT developments, their drawbacks, and the potential for future advances.

Chapter 12 discusses the use of artificial intelligence (AI) to make machines learn from the environment and make them capable of completing tasks, which helps to optimize their goals. AI, which has subfields such as machine learning, deep learning, and others, is interdisciplinary. Machine learning, which allows computers to automatically learn from their experience, may be achieved with computer programs that access and use them to understand. Deep learning is a subfield of machine learning, which processes or filters knowledge in the same way as the human brain. Here, to predict and classify the content, it uses a computer model that takes the input and filters it through various layers. These areas, such as artificial intelligence, machine learning, and deep learning, have made several developments in technology that have given the world a whole new dimension in each area.

Chapter 13 summarizes the important roles of certain AI-driven techniques (machine learning, deep learning, etc.) and AI-enabled imaging techniques for the study, prediction, and diagnosis of COVID-19 disease. Through social networking knowledge, the combined effort of powerful AI and image processing techniques can predict the initial trend of COVID-19 disease, identifying the most affected areas in each country, and predicting drug-protein interactions for the development of new drug vaccines. AI-empowered X-ray and computed tomography image acquisition and segmentation methods, however, help classify and diagnose patients minimally affected by COVID-19. This chapter also addresses an important set of open problems and future research concerns about AI-empowered COVID-19 handling procedures.

Chapter 14 mainly deals with the design of a machine learning model for the study of the transmission dynamics of COVID-19, a disease which is affecting the entire world. Ventilated patients with extreme acute respiratory distress being treated while quarantined in the ICU often face difficulties with their most basic human interactions, including communication, due to the respiratory disease, language issues or intubation. There are significant physical and psychological consequences to the inability of ICU patients to communicate. Researchers have created various types of software programs, such as Speech-Language Pathologist, in order to provide both health practitioners and caregivers with augmentative and complementary communication assistance.

SK Hafizul Islam Department of Computer Science and Engineering Indian Institute of Information Technology Kalyani West Bengal, India Email: hafi786@gmail.com

Debabrata Samanta Department of Computer Science CHRIST (Deemed to be University) Bengaluru, Karnataka Email: debabrata.samanta369@gmail.com May 2021