BIG DATA USAGE IN Ehealth and how it can be used in this field

BIG DATA USAGE IN Ehealth and how it can be used in this field. This an emerging technology and important one, since data is key

1. 2024 21st International Multi-Conference on Systems, Signals & Devices (SSD) School of Electrical and Information Technology, German Jordanian University, Amman, Jordan Ala’ Khalifeh BIG DATA IN HEALTHCARE: A REVIEW ON APPLICATIONS, TECHNOLOGIES, BENEFITS AND CHALLENGES 01 Princess Sumaya University for Technology, Amman, Jordan Raghda Bawaneh Princess Sumaya University for Technology, Amman, Jordan Ala Jallad

2. Agenda 02 Introducion Big data applications Big data techniques Types & Sources of big data Big data technologies Big data terminologies Conclusion & future work

3. 03 Introduction The healthcare sector is one of the biggest developing industries in the world and the accelerated patient cases worldwide result in the need to provide the best services at a minimum cost. Big data becomes crucial to provide professional patient care by analyzing and managing enormous health data, therefore, innovative big data tools are essential to ensure the management of health data

4. Using big data efficiently involves passing through multiple stages of collecting data from different departments, storing them, knowing exactly how to manage these data, then applying different methods 04

5. These features work in parallel with the known ”6Vs” of big data characteristics, Value, Volume, Velocity, Veracity, Variety, and Variability. Hesitation, Immortality and timeline, Privacy, Incompleteness, and Monarchy are the features that distinguishing the healthcare data. 05 1) Hesitation: Structured and unstructured data are used abundantly in healthcare industries, analyzing these data requires special tools, technologies, and a standardized framework that addressed the loose data by using regression and prediction/estimation methods 2) Immortality and timeline: the electronic health records in the health information system needs time especially for medical images such as MRI, CT, SPECT, and ECG. Which have significant sizes and may require extra uploading/downloading time. 3) Privacy: confidentiality of patients’ data is the key performance of the quality level in the healthcare sectors. 4) Incompleteness: is one of the big data challenges which occurs within these three types: Missing completely at random (MCAR), which occurs when the data loss is caused on out of control factors that are not related to the data itself. Missing at random (MAR), which happens when data loss can be deduced by factors you are aware of. Missing not at random (MNAR), which occurs when data loss is always related to the unobserved data. 5) Monarchy: patients have their medical data and they have the right to full access to medical data for patients by providing a separate account for them to store, manage, and share the designated data. Back to Agenda Page

6. Big Data Types Data Sources Medicine and Clinics Electronic Health Records & Medical Records (EHR/EMR) Medical Images Personal Health Record Echo Vital signs Radiography, magnetic resonance imaging (MRI), computer tomography (CT), and associated reports, patient identifications, treatments, pharmaceutical prescriptions, anesthesia activities, medical examination, diagnosis process, physician’s notes, and sensors’ records. Dexa scan, radiography, X-ray, CT, histology, positron-emission tomography (PET), MRI, elastography, ultrasound, tactile imaging, photoacoustic imaging, echocardiography, angiography. Allergies and drug’s reaction, chronic diseases, family background, illnesses, imaging reports, laboratory test results, medications, prescription records, surgeries and other procedures, vaccinations, and observations of daily living, and observations reported by patients. all Electrocardiograph (ECG) types. Temperature, pulse, respiratory rate, and blood pressure. Types & Sources of big data in healthcare sector 08

7. Big Data Types Data Sources Public Health Molecular data Molecular biology experiment Genomics, transcriptomics – whole genome sequencing, RNA seq, metabolomics –Nuclear Magnetic Resonance (NMR), mass spectrometry, proteomics – mass spectrometry, methylomics – pyrosequencing, and Chip-on-chip. Molecular cloning, polymerase chain reaction (PCR), macromolecule blotting and probing, microarrays, and next-generation sequencing. 09 Medical Experiments Human Parts Clinical trials Cells, tissues, and organs. Drug efficacy, toxicity, new treatment devices, and procedures. Medical literatures Journals/ articles knowledge PubMed, NEJM, the Lancet. Data MeSH. Back to Agenda Page

8. Back to Agenda Page Big data terminologies 10 Features knowledge discovery Data mining Big data analytical approach

9. Big data for education and research purposes One application in healthcare education is visual analytic, which is used in combining the exploitation technique and analysis of data, information representation and recognizing visual patterns Smart hospitals Big data and the application of IoT in hospitals can be used for a better circulation of the hospital materials. This can be done by implementing RFID technology into the healthcare organization Healthcare prediction and management Enormous application of big data in healthcare can be applied in diseases’ prediction, clinical outcomes, epidemic prediction, fraud detection, access medical data, provide faster way in retrieve the required data, propose the suitable design of medical devices, insurance transaction process, provide effective patient care, develop the pharmaceutical field, and medical treatment Big data applications in healthcare 11

10. Virtual assistants Virtual assistants such as Microsoft Cortana, Google Assistant, and Apple Siri, are supporting doctors in many ways. The assistant can help in converting a huge number of words into a single medical terminology by acting like an intermediate between the doctors and their patients Healthcare management and decision-making by using Google Maps and free public health data to prepare heat maps to find different issues such as the growth of chronic diseases within the population Big data applications in healthcare 12 Back to Agenda Page

11. Back to Agenda Page Techniques used in big data healthcare applications 13

12. Back to Agenda Page Technologies of big data 14 Hadoop MongoDB Elastic search RapidMiner Spark

13. Back to Agenda Page Technologies of big data 14 Hadoop The Apache Hadoop software [5] is an open-source Apache framework written in Java language, which allows different distributed processing for big data. However, this framework grants the distributed storage feature. So, Hadoop distributed file system (HDFS) needs to be pro- vided, then data are processed step-by-step using map-reduce algorithm, where data are divided into two subsets: training and testing.

14. Back to Agenda Page Technologies of big data 14 MongoDB: It is an NoSQL Document Database that has been recently used in storing big data. The main reason to use it is to facilitate the scaling method and to add more flexible features. By the usage of embedded files and arrays, MongoDB has many features, it is suitable to use in complex data, and different languages are designed to store data like images and videos and provides smooth access to a high volume of data. It is defined as easy-to- use technology due to its high performing capabilities.

15. Back to Agenda Page Technologies of big data 14 RapidMiner: It’s an analytical tool written in Java, used for data mining, suggesting, creating, and implementing prediction analysis. It’s an open-source and centralized software solution, [23], [24], that can be used in different languages to provide professional workflow.

16. Back to Agenda Page Technologies of big data 14 Elastic search Elastic search: The intelligent engine in elastic search is Lucene library [21], a software project, written in Java. Elastic search can find a relation between algorithms to match texts and store indexes [22]. This technology is characterized by many features such as scalability and provides a functional API,

17. Back to Agenda Page Technologies of big data 14 Spark A data processing tool widely used to execute data mining for large scale data, es- pecially for healthcare industries. Spark is clas - sified as one of the fastest technologies in data execution, also it has a large capacity to deal with streaming and large data types. Spark contains a vast number of libraries like Spark SQL, MLlib,Spark Streaming, GraphX [25]

18. Back to Agenda Page This paper outlines the most relevant big data sources, techniques and technologies used to process, store and analyze big data related to the healthcare sector. Predicting disease before it occurs based on patients’ electronic medical records (ERHs) considered one of the most useful outcomes of using big data. As a future work, a more expanded research work is currently under preparation that provides a more detailed analysis of relevant technologies and tools related to big data in healthcare industries which reflects positively on patient satisfaction. CONCLUSION AND FUTURE WORK 15

19. Thank you! Do you have any questions?