An Industrial IoT Approach for Pharmaceutical Industry Growth: Volume 2

By Valentina Emilia Balas, Vijender Kumar Solanki and Raghvendra Kumar

An Industrial IoT Approach for Pharmaceutical Industry Growth, Volume Two uses an innovative approach to explore how the Internet of Things (IoT) and big data can improve approaches and make discoveries. Rapid growth of the IoT has encouraged many companies in the manufacturing sector to make use of this technology to unlock its potential. Using clear language and real-world case studies, this book discusses systems level from both a human-factors point-of-view and the perspective of networking, databases, privacy and anti-spoofing. The wide variety in topics presented offers multiple perspectives on how to integrate the Internet of Things into pharmaceutical manufacturing.

This book represents a useful resource for researchers in pharmaceutical sciences, information and communication technologies, and those who specialize in healthcare and pharmacovigilance.

• Emphasizes efficiency in pharmaceutical manufacturing through an IoT/Big Data approach   
• Explores cutting-edge technologies through sensor enabled environments in the pharmaceutical industry
• Discusses system levels from both a human-factors point-of-view and the perspective of networking, databases, privacy and anti-spoofing

• Language: English
• Publisher: Academic Press
• Release date: May 15, 2020
• ISBN: 9780128213278

Book preview

IoT.

Chapter 1

Medical big data mining and processing in e-health care

A. Vidhyalakshmi¹ and C. Priya²,    ¹Department of Computer Science, Vels Institute of Science, Technology and Advanced Studies (VISTAS), Chennai, India,    ²Department of Information Technology, Vels Institute of Science, Technology and Advanced Studies (VISTAS), Chennai, India

Abstract

Health care is thought to be one of the business fields with the largest big data potential. Based on the prevailing definition, big data has a large amount of data which can be processed easily and can be modified or updated easily. These data can be quickly stored, processed, and transformed into valuable information using older technologies. At present, many new trends regarding new data resources and innovative data analysis are followed in medicine and health care. In practice, electronic health-care records, free open-source data, and the quantified self provide new approaches for analyzing data. Some of these advancements have been made in information extraction from the text data based on analytics, which is useful in data unlocking for analytics purposes from clinical documentation. Choosing big data approaches in the medicine and health-care fields has been lagging. This has led to the rise specific problems regarding data complexity and organizational, legal, and ethical challenges. With the growth of the uptake of big data in general, and medicine and health care in specific, innovative ideas and solutions are expected. Telemedicine is a new opportunity for the Internet of Things (IoT). This enables the specialist to consult a patient despite them being in different places. Medical image segmentation is needed for the analysis, storage, and protection of medical images in telemedicine. Telemedicine is defined by the World Health Organization (WHO) as the practice of medical care using interactive audiovisual and data communications. This includes the delivery of medical care services, diagnosis, consultation, treatment, as well as health education and the transfer of medical data. IoT-based applications mainly include remote patient monitoring and clinical monitoring. In addition, preventive measures-based applications are also part of smart health care. These applications require image processing-based technologies which could be integrated into medical health-care systems. Various types of input taken from cameras and processing of CT and MRI images could be integrated into IoT-based medical applications.

Keywords

Big data; IoT; health care; telemedicine; WHO; image processing

1.1 Introduction

In any industry big data can be changed by managing, analyzing, and leveraging the data. Health care is a promising area in which this could be applied to create a change in the field, which would result in many advantages. In health care, quality of life is improved by reducing treatment costs, epidemic outbreak prediction, and avoiding preventable diseases. The global population is increasing leading to new challenges for treatments and delivery methods. Health professionals are similar to business entrepreneurs, with a massive volume of data being obtained by them which could be selected and used by using the best strategies.

Any information related to human death, health conditions, quality of life, and reproductive outcomes is referred to as health data, whether for an individual or a population. Health data consist of clinical metrics along with environmental, socioeconomic, and health and wellness-based behavior information. The communication of individuals with health-care systems is collected as health data for processing and use by health-care providers. These data consist of a record of received services, service conditions of those data, and practical outcomes obtained in the clinic or information concerning those services [1]. By having this kind of framework most health data can be sourced. The advances in health information technology (IT) and eHealth care have been enhanced by the use of health data. The advancements in health care also enhance data security, data privacy, and ethical concerns. The main and major components of digital health are achieved by increasing data collection and usage.

Health data are split into two types: structured and unstructured. When the data are easily transferred between the health information system and in standardized form then the system is said to be structured [2]. The name of the patient, patient DOB, and output of a blood test report are stored in a structured data format whereas nonstandardized data are considered as unstructured health data [2]. Examples of unstructured data include emails, audio recordings, or physician notes about a patient. Based on the collection and use of health data, the information system is expanded, in a health-care field the data complexity is reduced with standardization [3]. In 2013, it was observed that 60% of data in the United States was unstructured [4].

The industrial internet is defined as the human services on the IoT, these terms suggest about expanding number of brilliant very quickly, then the gadgets will be moved among the connection between the gadgets and the amount of information and individuals. Some assessments have estimated that $120 billion was spent in 4 years for medicinal services in the IoT cloud, with human services IoT being the major part of this information in an unstructured format. This unstructured data information working within the Hadoop system has made Hadoop an interesting part of the progress into investigating a wide range of information. Recently, miscellaneous collections of screen gadgets are used, from glucose screens to fetal screens, fetal screens to electrocardiograms, and then to circulatory strains for each patient. Thereafter a subsequent visit from a doctor would be important with this huge amount of estimations.

When intelligent medical devices are able to interpret the results of other devices and gadgets, the requirements for direct doctor action may be replaced by a telephone call to the patient from a medical attendant. In addition patients can use devices at home and have the data transferred electronically. The possible results provided by human services IoT could result in immense improvements to patient care. In this human services substructure, utilization of the IoT can enable the installation of hardware, programming, sensors, and systems to permit these devices to communicate with one another and trade information.

The IoT has many possible cutting-edge innovations which could affect the whole range of businesses including health care. In today’s framework this can be considered as the interrelation of exceptionally identifiable items and gadgets with advantages and benefits which can be achieved by going beyond the machine-to-machine environment.

1.1.1 Types of big data

The definition and types of big data and are discussed next.

1.1.1.1 Structured

Data which can be stored, processed, and retrieved in a fixed form are known as structured data. These contain well-ordered information which can be promptly and perfectly stored. These data can be acquired from a database using search engine techniques. For example, the workers table in a firm’s database will be designed to include details of the employees, their positions, salaries, etc., in a well-structured manner.

1.1.1.2 Unstructured

Unstructured data are data that are unavailable in any specific form or structure. These kinds of data are complex to process and the processing time for these types of data can be excessive. Email is an example of unstructured data.

1.1.1.3 Semistructured

Data which contains both structured and unstructured formats is defined as semistructured data. Specifically this indicates information which has not been put under a particular database, and which consists of tags or vital information that separates individual elements of the data.

1.1.2 Characteristics of big data

1.1.2.1 Variety

Big data that are collected from many of multiple sources may be structured, unstructured, and/or semistructured. Previously, data may have been gathered from Excel sheets and databases, however nowadays the data are also sent in the forms emails, PDFs, photos, videos, audios, SM posts, and etc.

1.1.2.2 Velocity

Velocity is defined as the rate of speed at which data are generated in real time. In a broader context, it incorporates the rate of change, interconnection of incoming data sets at various speeds, and activity bursts.

1.1.2.3 Volume

The word big data has its meaning in the words themselves. Big data is purely a large amount of data that is being created on a day-by-day basis from various resources such as human interactions, social media platforms, machines, business processes, and networks. Such a large volume of information is saved in a data repository.

1.1.3 Integration of big data with medical imaging

Medical imaging and its processing plays a vital role in medicine in the United States, where about 600 million imaging procedures are performed annually. It is difficult manually to examine and store these images and also expensive and time consuming as hospitals need to protect and store them for several years in case of the need for future analysis by radiologists.

Medical imaging distributor care streams illustrate how often images were changed while analyzing the big data in health. The algorithms developed by physicians should analyze specific patterns in the hundreds of thousands of pixels in images and convert them into a numerical format for diagnosis. Furthermore, it could be feasible that radiologists in the future will no longer need to look at the images, but instead algorithms would analyze the outcomes as they are able to produce and remember a greater number of images. This will clearly affect the work of radiologists and their education and skills.

1.1.4 Advantages of health-care data management

The advantages of using health-care data management include the following:

• Producing 360 degree views of consumers, patients, and households, and deploying personalized, guided conversations by associating data from all available sources.

• Enhanced patient engagement with predictive modeling and analysis based on health-care data.

• Improved population health outcomes in specific geographic areas by tracking current health trends and predicting upcoming ones.

• Making informed, high-impact business decisions based on data insights.

• Understanding physician activity and aligning them with the organization’s goals.

• Predictive analysis is the biggest key benefit of big data. The analytics tool of big data can classify outputs correctly and, based on this, businesses and firms are able to make the correct decisions. Clearly it can help them to avoid and reduce risks and simultaneously optimize their operational efficiency.

• Analytics tools in big data can be used in social media platforms for data harnessing; businesses around the world are streamlining their digital marketing strategies to enhance the overall consumer experience. Big data provides insights into customer pain points and allows companies to improve upon their products and services.

• To be more specific, big data associates the most common information from a greater number of resources to create highly actionable insights. As much as 43% of organizations do not have the required tools to filter out noisy and unwanted data, therefore they need to spend a huge amount of money to catch the relevant data in bulk, which can be avoided by using big data tools, which save money and time.

• To develop and create more sales leads in companies, big data analytics is used to increase revenue. In businesses, big data analytics tools are used to know how effectively they are services and products are performing in the market and how their customers are responding. From this they can get an idea of where to invest their time and money more effectively.

• In big data insights, it is a must to always remain one step ahead of competitors. The market can be monitored to understand what types of promotions and offers are provided by competitors, to enable companies to decide on better offers for their customers. Also, big data insights give an idea into customer behavior in order to gain knowledge of customer needs and to provide an improved experience.

1.1.5 Challenges of health-care data management

The amount of health-care data available was expected to reach roughly 25,000 petabytes by 2020. Planning for and managing all that data can be an overwhelming, daunting task. Health-care organizations need to transition their operations toward a data-driven mentality: administrators and physicians must be diligent about collecting patient data, marketing departments must base their programs around data insights, and patients must be prompted to provide updated data whenever possible. Making data management a priority requires involvement from all players in the health-care industry—and this can present a challenge.

1.1.6 Health care as a big data database

Health care is a multidimensional process with the goal of prediction, prevention, and treating health problems in patients. The largest part of a health-care process is the health professionals, that is physicians and nurses, health facilities like clinics and hospitals that distribute medicines and use up-to-date treatment methodologies, and a supporting financing institution. Doctors can belong to different health sectors such as physiotherapy, midwifery, dentistry, nursing, psychology, and medicine. Based on an emergency situation health care is required at many levels. Doctors are the first point of consultation, that is for primary care, acute care requiring skilled professionals comes under secondary care, advanced medical investigation and treatment are called tertiary care, and highly uncommon diagnostic or surgical procedures are known as quaternary care. At all these levels, health professionals are responsible for specific types of information, such as patient’s medical history (diagnosis and prescriptions related data), medical and clinical data (like data from imaging and laboratory examinations), and other private or personal medical data. In the past, either handwritten notes or typed reports were used to store such medical records [5], even the outputs from medical tests were stored in a paper filing system. Obviously, this practice is now outdated, with the oldest case reports existing on a papyrus text from Egypt that dates back to 1600 BCE [6]. In Stanley Reiser’s words, the clinical case records freeze the episode of illness as a story in which patient, family and the doctor are a part of the plot [7].

With the development of computer systems health-care systems have digitized all clinical exams and medical records in a standard format in a widely adopted practice. In 2003, a division of the National Academies of Sciences, Engineering, and Medicine known as the Institute of Medicine coined the term electronic health records to represent record maintenance for the development of the health-care sector for the benefit of patients and clinicians.

1.1.7 Benefits of medical big data

The quality of health care is improved by using medical big data in an efficient manner, with this kind of technology used to increase analytical abilities, then predict epidemics, cure disease, build efficient health profiles, and enhance patients’ quality of life. Also, it plays a role in outcome improvement, preventable death avoidance, model prediction, and decreasing resource wastage. Big data is a method of understanding the biology of a disease by interlinking a huge collection of information to develop relational models with meaning. Medicinal big data is behind the drive for the process of developing models in therapy. We are able to know much more about a patient with this type of technology and share information as early as possible, by gathering the signs of severe diseases at an early stage for rapid and less expensive treatment. Even very small matters are considered in the medical field of big data analysis. The advantages of big data in the health-care domain include the following:

• Facility performance optimization;

• Reduction in energy cost;

• Ease of accessibility of information;

• Real-time updates and alerts;

• Proactive maintenance of equipment;

• Health-care delivery cost reduction;

• Reducing the costs of research and growth.

1.2 Architecture of big data in health care

A batch layer and streaming layer are useful to simultaneously receive medical data. When the data are saved in data nodes they are referred to as batch modes. In batch modes data are transferred to a semantic module which influences the meaning of the original data using an ontology store. Hereafter, data preprocessing like cleaning and filtering operations are performed to the output data before being processed. Then the extraction and selection of features is analyzed for the preprocessed data through various phases. Finally, the features of the data are used in classification models to predict future health condition of patients. This process works in an offline scenario. In the stream framework, information is collected from many other sources such as medical sensors connected to the patient’s body and when measuring blood pressure. Then, data synchronization is applied for the collected data according to the time and its missing