Blog

Copyright ©2017, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.DOI: 10.4018/978-1-5225-1002-4.ch006Chapter 699Diabetes PatientsMonitoring by CloudComputingABSTRACTThe healthcare system is important due to the focus on human care and the interference with human lives. In recent years, we have witnessed a rapid rise in e-healthcaretechnologies such as Electronic Health Records (EHR) and the importance ofemergency detection and response. Cloud computing is one of the new approachesin distributed systems that can handle some of the challenges of smart healthcarein terms of security, sharing, integration and management. In this study, an architecture design of a cloud-based pervasive healthcare system for diabetes treatmenthas been proposed. For this, three different components are defined as follows:(1) The home context manager which gathers necessary information from patientswhile simultaneously providing feedback, (2) a patient health record manager thatis accessible by nurses or physicians at the hospital, and (3) a diabetes managementsystem which is located with the cloud infra-structure for managing and accessingpatient’s information. The performance of proposed architecture is demonstratedthrough a user scenario.Sepideh PoorejbariPervasive and Cloud Computing Laboratory, IranHamed Vahdat-NejadUniversity of Birjand, IranWathiq MansoorUniversity in Dubai, UAEDiabetes Patients Monitoring by Cloud Computing100INTRODUCTIONInformation technology can play a vital role in healthcare services in terms of electronic health. Recent advances in e-health can be broadly defined as the applicationof information and communication technologies in healthcare systems (Varshney,2009). Making use of the internet for storing, accessing and modifying healthcareinformation and digitizing many processes and tasks is a necessary step for realizing e-health. In this case, we have the advantages of e-health such as a rise in thequality of services in aging societies, reduction in cost and in medical errors andthe ease by which data can be moved to the right place. However, digitizing paperbased records, collecting and storing medical information as well as lack of suitabletechnology for preventive care can become rather challenging.After the emergence of the pervasive computing paradigm, pervasive healthcaretechnology has been proposed to support a wide range of applications and servicesincluding patient monitoring and emergency response. However, they simultaneouslyintroduce several challenges including data storage and management, interoperability, availability of resources and ubiquitous access issues (Ziefle & Rocker, 2010).Diabetes is one of the major chronic diseases in the world. Diabetes, often referred to by doctors as diabetes mellitus, describes a group of metabolic diseasesin which the person has high blood glucose (blood sugar), either because insulinproduction is inadequate, or because the body’s cells do not respond properly toinsulin, or both. Diabetes manifests itself in three types:Type 1: This type of diabetes is usually diagnosed in children and young adults,and was previously known as juvenile diabetes. Only 5% of people with diabetes have this form of the disease. In this type of diabetes, the body does notproduce insulin.Type 2: Is a problem with your body that causes blood glucose (sugar) levels torise higher than normal. This is also called hyperglycemia. Type 2 is the mostcommon form of diabetes; About 90 percent of people with diabetes havetype 2 diabetes.Type 3: Gestational Diabetes is a temporary condition that occurs during pregnancy.It affects approximately 2 to 4 percent of all pregnancies and involves an increased risk of developing diabetes for both the mother and child.All forms of diabetes increase a patient’s risk of emerging different health complications. Short-term complications such as hypoglycemia and hyperglycemia (very lowand high blood glucose), and long-term complications such as eyes, heart, kidneys,nerves and feet failure are serious and life-threatening. The proper management ofblood glucose levels reduces the risk of developing these complications. FactorsDiabetes Patients Monitoring by Cloud Computing101such as the illness that patient suffers, treatment, physical and psychological stress,physical activity, drugs and diet (meal plan) can cause unpredictable and dangerous consequences such as hypoglycemia, hyperglycemia, and falling into a coma.The management of diabetes is becoming an increasingly important problemworldwide. In 2014, according to the International Diabetes Federation, at least 387million people (or 8.3% world population) suffered from diabetes and it is expectedthat by 2035, the number of diabetes will increase to more than 590 million.Case Study: This scenario originates from the Imam Reza hospital, one of the leading and technical hospitals in Mashhad, Iran.Mr. Toosi, a 62 year old man has suffered from type 2 diabetes for more than 10years. In this period of time, he has caught different health complications such asdiabetes foot, eye problems, high blood pressure, and heart problems. His prescription includes 20 units of insulin per day, 12 units in the morning and 8 units atnight. In addition he has to check his glucose level on a daily basis. On one earlymorning, Mr. Toosi woke up with shortness of breath and weakness in his body.Without special attention, he had his breakfast and injects his insulin. After two hourshe feels more pain and weakness in his chest and body, but because he is alone athome he prefers to wait until one of the family members returns home, before heading out for a check-up. When Mr Toosi’s son arrives home, he immediately driveshis father to the nearest hospital. At the hospital, the physicians and nurses runsome treatments, however due to the lack of medical information and patient healthrecords they are unable to make precise decisions and prefer to send the patientto the hospital where Mr. Toosi’s main physician is available. Unfortunately, aftersending the patient to another hospital he falls into a diabetes coma and passesaway after a second heart attack.The above scenario leads to a few issues that need to be addressed in pervasivediabetes health system. Here we focus on the following two problems:• Monitoring patient remotely at home and detecting and managing differentsituations.• Accessing patient health records and medical history at anytime and anywhere by legal persons.The concept of cloud-based pervasive healthcare system is a new paradigm forthe healthcare sector that uses cloud computing to treat, manage and control patientspervasively. The systems are supported by different algorithms, cloud infrastructures,smart homes, devices, and sensors and create several service types according to theircontext and environment. This paper presents a Cloud-based pervasive healthcareDiabetes Patients Monitoring by Cloud Computing102system architecture for treating diabetes, in order to manage and control diabeticpatients and reduce the hypoglycemia and hyperglycemia conditions and consequentlytheir risks. The system, which we will refer to as DICPer-Health, is designed tocontrol patients pervasively at their homes. Three environments including a home,hospital and cloud structure are considered as main parts of this system, by whicheach of these sections has its own components and acts separately.We have selected diabetes type 2 as the chronic condition cannot be cured andis the condition of 9/10 diabetes patients. In type 2 diabetes, the body cannot useinsulin properly. This is called insulin resistance. Type 2 is treated with lifestylechanges, oral medications, and insulin injections. This type of diabetes usuallygets worse over time, and in order for individuals with type 2 diabetes to controltheir blood glucose levels, they need to eat healthy, stay active, and use prescribeddrugs appropriately. In this case we have considered three important factors in ourframework; diet, activity, and insulin. These three factors control and manage thelifestyle of type 2 diabetes patients.The following sections of the paper are organized as follows; Section (2) presentsthe pervasive healthcare projects and related works. Section (3) presents architectureto support diabetes treatment. The architecture parts are described in subsections.Section (4) presents the evaluation of work according to real life patient scenario.Finally, section 5 will conclude the paper.RELATED WORKSToday, there is a great amount of research work in the field of pervasive healthcareto improve e-health services; however only a few number target the use of cloudinfrastructure as a new IT paradigm and are surveyed as below.“The Integrated Cloud-based Healthcare Infrastructure” project, ICHI has beendeveloped in Edinburgh Napier University of United Kingdom. ICHI presents a systemthat integrates a formal care system (DACAR) with an informal care system (MicrosoftHealth Vault) that enables not only sharing and access of health records right alongthe patient pathway, but also provides a high level of security and privacy withina cloud environment (Ekonomou, Fan, Buchanan, & Thuemmler, 2011). Anotherproject in the University of Central Greece, “Bringing IoT and Cloud Computingtowards Pervasive Healthcare”, IoTC, proposes a platform based on cloud computing for management of mobile and wearable healthcare sensors (Mu-Hsing Kuo,2011). In another research project concluded at the University of Greece, namely“Managing Wearable Sensor Data through Cloud Computing” (MWSC), a wearabletextile platform that collects motion and heartbeat data and stores them wirelesslyDiabetes Patients Monitoring by Cloud Computing103on an open cloud infrastructure for monitoring and further processing was studied(Wan, Zou, Ullah, Lai, Zhou & Wang, 2013). “Cloud-Enabl

ed Wireless Body AreaNetworks for Pervasive Healthcare (CWBAN)” is another article in the same contextthat focuses on a cloud-enabled WBAN architecture and its applications are withinpervasive healthcare systems. CWBAN develops WBANs with MCC (Mobile CloudComputing) capability, a Cloud-Enabled WBAN (Woon Ahn, Cheng, Baek, Jo &Chen, 2013). This project has also been developed in different universities includingSouth China University of Technology, King Saud University, and the University ofBritish Columbia. Another article namely “An Auto-Scaling Mechanism for VirtualResources to Support Mobile, Pervasive, and Real-Time Healthcare Applications inCloud Computing” (RTHA) proposes a novel server-side auto-scaling mechanism.The model is based on cloud computing with virtualization technologies in collaboration with the University of Houston and Korea University (Corredor, Tarrio,Bernardos, & Casar, 2013).As mentioned above, the number of pervasive healthcare projects related tocloud computing for diabetes are few. Moving forward, we have scrutinized just twoarticles that introduce a personal health system (PHS) to manage diabetic patients.One of these projects at the University of London, is “COMMODITY12: A Smarte-Health Environment for Diabetes Management” that emphasizes on designingthe PHS to address major problems of both diabetic patients and doctors who treatdiabetes (Kafah et al., 2013). COMMODITY12 consists of ambient, wearable andportable devices, which acquire, monitor and communicate physiological parametersand other health factors and vital body signals of a patient. In this system, there areintelligent agents that use expert biomedical knowledge to interpret data and thenpresent a feedback from a patient’s health status directly to the patient from thedevice (Kafah et al., 2013).University of Murcia, Spain, has developed another healthcare project for diabetes,“An Internet of Things-based Personal device for Diabetes Therapy Managementin Ambient Assisted Living (AAL)”, that presents a personal diabetes managementdevice based on the Internet of Things. The target is to support a patient’s insulintherapy to decrease hyperglycemia and hypoglycemia counts and the risks involved(Jara, Zamora, & Skarmeta, 2011). This project focuses more on insulin dosagebased on mobile assistance services. The project considers different factors such asthe illness that patients suffer, drugs, treatments, stress, physical activity and meal(diet) for insulin therapy.One related survey is “An Introduction to Cloud-Based Pervasive HealthcareSystems”, that reviews different projects in healthcare sector with the focus on cloudcomputing (Poorejbari & Vahdat-nejad, 2014).Diabetes Patients Monitoring by Cloud Computing104THE PROPOSED FRAMEWORKOverviewTo introduce DICPer-Health, our first initiative is to present the general architecture of the system in order to briefly describe the main components. We will thenconcentrate on the description of three key components: Home Context Manager,Hospital Environment and Cloud Infrastructure.Figure 1 depicts our proposed architecture for a smart healthcare environment.The figure consists of three main components which are connected via Internet:(i) The home context manager which gathers necessary information from patientswhile simultaneously providing feedback, (ii) a patient health record form that isaccessible by nurses or physicians at the hospital, and (iii) a diabetes managementsystem which is located with the cloud infrastructure for managing and accessingpatient’s information.In our proposed architecture, the home context manager plays a major role incollecting, storing and processing data. Once the initial procedures are complete,Figure 1. General Architecture of DICPer-HealthDiabetes Patients Monitoring by Cloud Computing105the system presents alerts, suitable treatments, and advices to the patient. The function of the home context manager and its components will be described in a latersection.Home Context ManagerThe main role of the home context manager is collecting different information suchas blood glucose level, blood pressure, and heart rates from patients through varioussensors at specific times. After gathering patient health parameters, all the requiredinformation is stored in a context database. The inference engine then infers twodifferent important conditions; High risk and emergency situations. According tothe conditions and patient health history, suitable advices and alarms will be communicated to the patient’s smart device or in an application connected to nurses orexperts. The home context manager consists of the following components.Wireless SensorsSensors are essential components in smart environments which sense and collectphysical parameters. We utilize three types of wireless sensors in our framework,which gather fundamental physiological information from diabetes. The sensorsare depicted in Table 1 and described below with the objective of sensing our requirements.• Glucose monitoring by finger prick (GlucoTel Sensor).• Blood pressure monitoring (PressureTel Sensor)• Heart rate monitoring (Pulse Sensor)None of the above sensors are expected to be considered obtrusive by the patients.Patients without any special knowledge can use them with ease.Table 1. Schema of context table ParameterDateTime8 A.M.10 A.M.3 P.M.10 P.M.Blood GlucoseBlood PressureHeart Rate Diabetes Patients Monitoring by Cloud Computing106Context CollectorAfter sensing raw data by sensors, the context collector collects the data and deliversthe information to the context database. 6LoWPAN is a technology for transferringsensed data to smart devices. This technology connects wireless clinical deviceslike glucometer to the smart environments or gateways. We proposed 6LoWPANtechnology for sending data from the sensors to the context collectors and then byusing SQL statements, context collector inserts data into the context database.Context DatabaseThis database consists of one table, namely the context table by which physiologicalparameters are the defined fields (Table 1).The parameter field is designed to evaluate a patients important physiologicalfactors such as blood glucose, blood pressure, and heart rate levels. The date andtime fields reveal patient check-up timings. For diabetic patients monitoring andmanaging blood glucose is an essential task as blood glucose levels should remainwithin normal ranges. There are two types of blood glucose measures that are veryimportant in a diabetic treatment; FBS1 and the THG2 level check-up. In this casewe have proposed four specific times in order to measure these parameters. It beginsearly morning at 8 A.M. for an initial measuring FBS, 10 A.M. for measuring the 2hours glucose level after breakfast meal, 3 P.M. for evaluating the 2 hours glucoselevel after lunch time, and 10 P.M. for evaluating the 2 hours glucose level afterdinner. In addition the patient’s blood pressure and heart rate levels will be checkedsimultaneously. After monitoring all the parameters via wireless sensors, the senseddata will be stored in a context table, by which the inference engine can use themfor inferring suitable outcomes.Inference EngineThe inference engine determines two principal situations; high risk and emergencybased upon specified rules and algorithms. The inputs of the inference engine arethe context table data and, the outputs of this function in high risk condition, is apatient form that shows alarms and useful advices such as practical diet, activityand correct insulin dosage, and in emergency situation, is a nurse or physician formthat presents an alarm and dangerous factors to the nurse or experts.In our proposed framework, the data interpretation is based on the specific rulesand defaults. Moving forward, we will describe the basic definitions and default rules.Diabetes Patients Monitoring by Cloud Computing107Table 2 presents the normal and abnormal range of blood glucose levels and theunit of blood glucose is measured as mg/dl (milligrams per deciliter). And all physicians and experts try to keep the patient blood glucose level at the normal range.So, the rules of our proposed architecture are defined according to these rangesfor distinguishing hypoglycemia and hyperglycemia conditions.The other important parameter is the blood pressure levels that should be considered. Most diabetic patients have a high blood pressure problem that requiresmonitoring. The normal range of blood pressure is between 80-120 mmHg (millimeter of mercury), ranges are considered abnormal when reaching the upper rangeat approximately 135 mmHg. If the blood pressure of a patient becomes more than135 mmHg, then the system recommends some treatments for this condition.Heart rate is another essential factor in diabetes treatment. The normal range ofa heart pulse is 60-100 beats per minute (bpm). In our system, we have labeled itas “Low”, if the number of the heart rate becomes less than 50 bpm, and “High” ifthe number of the heart pulse reaches more than 100 bpm.Diet and meals play an essential role in diabetes treatment and control of bloodglucose level. According to nutritional views, diabetes patients are required to havea specific amount of nutrients such as carbohydrate, fiber, sodium and protein intheir daily meals. The proposed amounts of nutrients are classified in Table 3.For daily activities, the body needs energy and this energy comes from foods.Each food has a specific amount of calories. The number of calories the body consumes in a day is different for every individual.In our proposed framework, according to the calculated energy for each patient,the amount of carbohydrate that should be used by patients is at least 55 percentand at most 70 percent of a patient total energy. The amount of fiber is 20-30 grams,Table 2. Glucose different ranges Very LowLowNormalHighVery HighGlucose Level