Big data in healthcare

Big data in healthcare Cite as: AIP Conference Proceedings 2424, 050005 (2022); https://doi.org/10.1063/5.0077737 Published Online: 21 March 2022 Siddhant Rajkumar and Yasha Hasija AIP Conference Proceedings 2424, 050005 (2022); https://doi.org/10.1063/5.0077737 © 2022 Author(s). 2424, 050005 Big Data in Healthcare Siddhant Rajkumara) and Yasha Hasijab) Delhi Technological University, Shahbad Daulatpur, Main Bawana Road, Delhi-110042, India a) Corresponding Author: [email protected] b) [email protected] Abstract. Big data is the large amount of data that has been popular for the past few decades because of its many uses. Big data is already being used in fields such as business management and machine learning. In recent years it is storming its way in the healthcare industry as well. Electronic health records, efficient staffing, supply chain management is some of the big data applications in the healthcare industry. These applications are discussed in detail later in the paper. Over the years, the need for big data has been increasing because of various factors such as improving patient outcomes, efficiently managing healthcare-related data such as medical records, past diagnostic reports and prescriptions and, documents of various medical tests. Big data can help improve the patient's overall care while keeping the treatment cost low as there would be no need to run redundant tests. But many factors are restricting the use of big data in the field of healthcare. These could be the incompatibility of software or the unwillingness of organizations to share data. But over the years, because of big data, the healthcare industry has been improving towards developing new analytical and computational software that could revolutionize the healthcare industry. Keywords—Big data, business management, Machine learning, healthcare, Electronic healthcare records, Supply chain management. INTRODUCTION Big data signifies a large amount of data that is unmanageable via regular software and online systems. It requires an unprecedented amount of storage, computational, and analytical power 1. As per Douglas Laney's definition, data grows in three aspects which is called the “3 V’s of big data” 2. FIGURE 1. 3 V’s of big data as per Douglas Laney In big data, the big part represents the huge volume of the data. Velocity means the flow at which data is collected and is made more accessible for further research. And variety shows the different types of data that the firm can collect, be it organized or unorganized. Since big data is undecipherable using traditional and simple software, we need to use high-end cost-effective systems with sufficient computational power to complete such a task. Such systems generally use artificial intelligence and various novel fusion algorithms3. Without these, it would be impossible to decipher such a large amount of unorganized data. Along with appropriate computational power, we also require proper hardware to support the automated decisionmaking process. Lack of adequate hardware can lead to big data being hazy. As long as we have good analytical tools, Proceedings of the International Conference on Computational Intelligence and Computing Applications -21 (ICCICA-21) AIP Conf. Proc. 2424, 050005-1–050005-8; https://doi.org/10.1063/5.0077737 Published by AIP Publishing. 978-0-7354-4179-8/$30.00 050005-1 we can draw meaningful insights from big data, making social services like healthcare and public security, both physical and cyber, more interactive and efficient4. For big data management, experts from various backgrounds such as biology, computer science, mathematics, and statistics collaborate to manage the data successfully. The large amount of data collected using multiple wearable’s are shared on cloud storage. Cloud storage has pre-installed software, which helps in data mining and makes it easily accessible. These analytical methods convert data into knowledge. Finally, this knowledge is displayed in graphical form by graphic designers. Big data has brought about revolutionary changes in today's world. These changes are not limited to any particular field but throughout all industries. Big data has brought many changes in the way we manage and analyze the data. Sharing of data between different departments has also increased therefore increasing collaboration. These changes are especially prominent and beneficial in the healthcare industry. In healthcare, big data analysis has brought about significant changes. These changes can help reduce the treatment cost as there is no need to perform repetitive tests5. The information regarding the previous diagnostics and prescriptions is also available. Information about the patient, like allergies, is also available to prevent accidentally prescribing unsuitable drugs. Diagnostics regarding particular symptoms are already present, making it possible to avoid and stabilize preventable diseases. Big data can also help us identify and prevent the outbreak of an epidemic 5. In general, it increases the quality of life. Throughout the world, as healthcare facilities improve, the average human life span also increases. The increasing population causes more problems in current treatment delivery methods. Thus, a massive amount of data reservoir helps the healthcare professionals to utilize the available resources and personnel efficiently treat all the patients. TABLE 1. Some of the major startups in healthcare that are using big data6 Sno. 1. Name Komodo Health Country USA 2. Flatiron Health USA 3. Evidation health USA 4. Sema4 USA 5. EXscientia UK 6. iCarbonX China 7. Insilico Medicine Russia 8. PurpleDocs India Product Platform that uses real-time data to improve medical decisions. Gather data on cancer treatment from various sources and analyses it. They also connect oncologists, medical researchers and various health related organizations that work in the field of cancer. It provides an extremely fast and easy to use platform that connects individuals and health industry. They are developing a medical data analysis platform that provides personalized care to the patients. They are using Artificial Intelligence and big data analysis to speed up drug discovery process. They are using medical data to create a digital ecosystem based on individual data. They are trying to use artificial intelligence and big data analysis to increase longevity of human life span. They are trying to make healthcare data more accessible and secure. BIG DATA FROM HEALTHCARE INDUSTRY In the Healthcare industry data can be obtained from a variety sources like government agencies, patient portals, Electronic Healthcare Records, Research databases, genetic databases, Public records, search engines, smart phones payer records and wearable sensors. The below image show various sources of data in the healthcare industry. 050005-2 FIGURE 2. Different sources of data in Healthcare industry NEED FOR BIG DATA IN HEALTHCARE In many countries, the prices of healthcare are drastically increasing 10. Thus it is clear that we need an intelligent approach that would make healthcare more affordable to people. Earlier, insurance companies had a pay-for-service method to pay the hospital for their service regardless of the patient's outcome. But now, they are moving towards plans that prioritize the patient's condition. Thus data sharing has also increased. Previously health workers had no reason to share their patient's information with other organizations, hospitals, or research centers, but now they do. Sharing data will help maintain a shared pool of data that will enable researchers to draw more valuable insights and improve healthcare facilities. More amount of data will be beneficial for developing analytical patterns and trends. Better data analysis will also help improve the overall patient care offered while keeping the cost minimum 28. Finally, the current healthcare system is becoming more data-driven as doctors rely more on evidence and past data than their schooling and professional opinion. Thus, collecting a large amount of data is becoming more and more important, and we need big data analytics to manage this data. HOW BIG DATA HELPS THE PATIENTS Big data can be implemented in various sectors of the healthcare industry. Some of the common examples of using big data in the healthcare industry are in staffing, Electronic Healthcare Records, Real-time alerting, Enhancing patient care, Innovation and development of the healthcare industry, Medical Imaging, Predictive Healthcare Analysis, Supply Chain Management, Tele-medicine and Suicide Prediction. Based on past data we can predict the number of patients that will be visiting at a particular time and based on that determine the number staff that is required. This can help us solve the problem of over as well as under staffing. Electronic Healthcare records are a collection of large amount of medical data that contains the medical history of the patient along with past medical test reports and diagnostics. We can also use wearable sensors for Real-time alerting in case of high-risk patients. Tele-medicine is also an important aspect of the Healthcare industry. Here the doctors are patient are not in the same room thus the doctor operated from a remote location using advanced technologies such as video conferencing and robotics. INCREASE IN USE OF BIG DATA Over the years big data has experienced a massive growth especially in the healthcare industry. This growth can easily be identified with the help of the number of research papers published over the years. As the healthcare industry in becoming more and more data driven, large amount of data is generated every day. Thus to combat this increase in data appropriate research and advancements are also required. The graph below shows the increase in publications regarding big data in healthcare over the years. 050005-3 FIGURE 3. Increase In Healthcare Publications Over The Years MEDICAL IMAGING AND BIG DATA Over the years medical imaging has become a major part of the healthcare industry. These images provide the healthcare professionals with a more elaborate and accurate view of the patients conditions, thus in order to store all this data we use of big data. Now, medical images are being converted into quantitative data which enables faster and easier analysis and diagnosis. Various imaging software and computational software are being used are being used for processing and analysis of medical images and different types of scans. The table below shows the different types of scans that are supported by different software. FIGURE 4. Different Types Of Medical Scans Supported By Different Processing Software PERSONALIZED TREATMENT USING BIG DATA Using various computational methods in combination with big data we can come close to personalized treatment i.e. different treatment for every individual which is most suitable for them. In my opinion we can analyze the 3D structure of the genome to design personalized drugs for the patients. This will not only help in reducing side effects 050005-4 but also improve the patient care offered to the patient. Currently it is very difficult to analyze the complete structure of the genome. But we can proceed towards the development of personalized treatment by analyzing a single target protein and using molecular dynamics and molecular simulations to identify the most suitable drug for inhibiting the target site. Process Visualize the 3D Structure of the Protein We can use software called VMD (Visual Molecular Dynamics) for analysing the 3D structure of the protein in various conformations to identify the appropriate target site. VMD also provides with various filters and setting for better analysis. FIGURE 5. Shows the 3D representation of protein UDP-N acetylmuramoyl L-alanine:D glutamate ligase. Identification of Binding Site, Molecular Docking and Dynamic Simulations We can identify the binding site for various proteins using different online tools such as CASTp, for molecular docking and dynamic simulations. We can later perform molecular docking for the desired drug to figure the position that has least binding energy i.e. shows best and most stable binding. For this we can use software such as AutoDock, PyRx Visualizer etc. Next step is to perform dynamic simulations for the most stable conformation to check the effectiveness of the drug. For this we can use GROMACS and DESMOND. Molecular Docking follows a Lamarckian Genetic Algorithm. In this algorithm the first step is to create a random population of individuals. Here the size of the population is determined by the user. Each individual has 3 genes x, y, and z. for uniform distribution of the population. The fitness of the population is also very important. In many cases there is an inverse mapping function to identify and yield a genotype from a particular phenotype, thus we can finish a local search by replacing the individuals with the result. 050005-5 FIGURE 6. Shows the genotyoe phenotype search How %ig 'ata &an +elp in Personalised 7reatment With the help of Big data we store all the findings from various research papers regarding the analysis of a particular protein, its corresponding disease and the drugs required for treatment. This would greatly reduce the time required for identification of binding site and required drug. Various past research results will also be easily available for molecular docking and dynamic simulations. Based on past medical records also figure out the most suitable drug for every single individual. Thus over time we will have an abundant amount of data to analyse and predict the structure and conformation of the drug for every single individual which is best suitable for them. OBSTACLES TOWARDS USING BIG DATA One of the most significant concerns regarding implementation of big data in the field of healthcare is that medical data spread about various sources, follow different rules and regulations regarding data sharing, and use varying software to store the data29. Therefore it is necessary to implement standard guidelines and platforms to store data. With this, it would be easier for various organizations to collaborate. It is also important to utilize new and upcoming data sharing and management methods that other industries are already using. FIGURE 6. Obstacles in implementing Big data in Healthcare. 050005-6 CONCLUSION Nowadays, a large amount of data is generated in the medical industry using genome analysis, biometric analysis, and various sensors. It is required to store all this data and analyze it. This data study can give us significant insight into medical techniques and treatments, thus improving patient care. Now that big data is increasing rapidly in the medical industry, its pros and cons are beginning to surface. Big data help improve the medical care offered to the patient and can be used to maintain proper records of patients undergoing treatment. However, big data is challenging to keep and share because different states and countries have unique guidelines. Despite this, the advantages of big data outweigh the disadvantages. Thus it is also called the revolutionary step in development in the healthcare industry. Over the years, many new insights and discoveries have been made using big data, such as using antidepressants to cure certain lung cancer types. Using predictive and descriptive data analysis, we can ensure the proper functioning of hospitals. We can achieve this by adequate staffing and supply management. In the future, there is a possibility of developing computational systems capable of predicting a drug's response on a patient using the patient's genome sequence. These computational systems would significantly reduce the likelihood of side effects from taking medicine. ACKNOWLEDGMENTS My initial thank is addressed to my mentor Dr. Yasha Hasija, Associate Professor, Department of Biotechnology, Delhi Technological University, who gave me this opportunity to work in a project under her. It was her enigmatic supervision, constant encouragement and expert guidance which have enabled me to complete this work. I humbly seize this opportunity to express my gratitude to her. I would also like to extend my sincere gratitude to Professor Pravir Kumar for providing infrastructure and facilities. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. S. Dash, S.K. Shakyawar, M. Sharma, and S. Kaushik, J. Big Data 6, 54 (2019) D. Laney, META Gr. Res. Note 6, 1 (2001). A. De Mauro, M. Greco, and M. Grimaldi, Libr. Rev. (2016). J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, Futur. Gener. Comput. Syst. 29, 1645 (2013). Healthcare Analytics Software - Discover Healthcare BI Tools. Accessed April 14, 2021. https://www.datapine.com/healthcare-analytics Top 60 Big Data startups in Healthcare. Accessed April 14, 2021. https://medicalstartups.org/top/bigdata/ D.C.C. AOCNP, Clin. J. Oncol. Nurs. 19, 153 (2015). M. Reisman, Pharm. Ther. 42, 572 (2017). K. Ambert, S. Beaune, A. Paré, A. Chaibi, L. Briard, A. Bhattacharjee, V. Bharadwaj, K. Sumanth, and K. Crowe, Trusted Analytics Platform (TAP) Is a Collaborative Environment for Creating Advanced Analytics and Applications to Help Hospitals Improve Patient Care and Resource Allocation. French Hospital Uses Trusted Analytics Platform to Predict Emergency Department Visits and Hospital Admissions The big-data revolution in US health care: Accelerating value and innovation | McKinsey. Accessed April 14, 2021. https://www.mckinsey.com/industries/healthcare-systems-and-services/our-insights/the-big-datarevolution-in-us-health-care# K. Shameer, M.A. Badgeley, R. Miotto, B.S. Glicksberg, J.W. Morgan, and J.T. Dudley, Brief. Bioinform. 18, 105 (2017). L.H. Mellett and G. Bousquet, Circulation 127, (2013). How Big Data Helps To Tackle The No 1 Cause Of Accidental Death In the U.S. Accessed April 14, 2021. https://www.forbes.com/sites/bernardmarr/2017/01/16/how-big-data-helps-to-tackle-the-no-1-cause-ofaccidental-death-in-the-u-s/?sh=6a96796939ca J.H. Hibbard and E. Peters, Annu. Rev. Public Health 24, 413 (2003). A. Zingone, D. Brown, E.D. Bowman, O.M. Vidal, J. Sage, J. Neal, and B.M. Ryan, Cancer Treat. Res. Commun. 10, 33 (2017). Cancer Cures Could Already Exist In Big Data. Accessed April 14, 2021. https://www.fastcompany.com/3063530/cancer-cures-could-already-exist-in-big-data See Top 10 Analytics & Business Intelligence Trends For 2021. Accessed April 14, 2021. https://www.datapine.com/blog/business-intelligence-trends/ 050005-7 18. Online Business Intelligence Tools - See Affordable BI Software. Accessed April 14, 2021. https://www.datapine.com/online-bi-tools 19. R. Wootton, Bmj 323, 557 (2001). 20. N.M. Hjelm, J. Telemed. Telecare 11, 60 (2005). 21. S. Loue and S. Loue, in Leg. Issues Soc. Work Pract. Res. (Springer International Publishing, 2018), pp. 183– 206. 22. M.J. Yaffe, in Semin. Nucl. Med. (Elsevier, 2019), pp. 94–104.. 23. How Big Data Shaping The Supply Chain. Accessed May 19, 2021. https://www.slideshare.net/HafizullahMohdAmin/group-assignment-how-big-data-shaping-the-supply-chain 24. S. Alotaibi and R. Mehmood, in Int. Conf. Smart Cities, Infrastructure, Technol. Appl. (Springer, 2017), pp. 207– 215. 25. M.A. Morris, B. Saboury, B. Burkett, J. Gao, and E.L. Siegel, J. Thorac. Imaging 33, 4 (2018). 26. A. Tahmassebi, A. Ehtemami, B. Mohebali, A.H. Gandomi, K. Pinker, and A. Meyer-Baese, in (SPIE-Intl Soc Optical Eng, 2019), p. 13. 27. Top 10 Data Visualization Techniques, Concepts & Methods In Business. Accessed April 14, 2021. https://www.datapine.com/blog/data-visualization-techniques-concepts-and-methods/ 28. B. Feldman, E.M. Martin, and T. Skotnes, Big Data in Healthcare Hype and Hope (2012) 29. J. Adler‐milstein and E. Pfeifer, Milbank Q. 95, 117 (2017). 050005-8