Documente Academic
Documente Profesional
Documente Cultură
Abstract
Health care sector has grown tremendously in stored in hard copy form, the current trend is
last few decades. The health care sector has towards the rapid digitization of these large
generated huge amounts of data that has huge amounts of data. There are different types of
volume, enormous velocity and vast variety. data sources which generates these enormous
Also it comes from a variety of new sources as amounts of data. Big data in healthcare refers to
hospitals are now tend to implemented electronic electronic health care records (EHR) that is quite
health record (EHR) systems. These sources large and complex that they are difficult to
have strained the existing capabilities of existing manage with traditional software and/or
conventional relational database management hardware. Also, they are not easily managed with
systems. In such scenario, big data solutions traditional or common data management tools
offer to harness these massive, heterogeneous and methods. Using the technologies that able to
and complex data sets to obtain more meaningful deal with such Big Data will offer many
This paper basically discusses the impact of Healthcare big data refers to the vast quantities
implementing the big data solutions on the of data that is now available to healthcare
challenges and their solutions and available healthcare information and the rise of value-
platform and tools to implement Big data based care, the industry has taken advantage of
analytics in health care sector. big data and analytics to make strategic business
decisions. Faced with the challenges of
Keywords: Big Data, Health Care, Big Data healthcare data volume, velocity, variety, and
Analytics, oppurtunities and challenges. veracity, health systems need to adopt
technology capable of collecting, storing, and
I. Introduction
analyzing this information to produce actionable
The health care sector grows rapidly in last 30
insight.
years. The healthcare industry historically has
generated large amounts of data, driven by Particularly, big data analytics in healthcare
record keeping, compliance & regulatory enables analysis of the large datasets from
requirements and patient care. While most data is thousands of patients, identifying clusters and
1
correlation between datasets, as well as and unstructured. During earlier days,
developing predictive models using data mining spreadsheets and databases were the only sources
techniques. Big data analytics in medicine and of data considered by most of the applications.
healthcare integrates analysis of several scientific Nowadays, data in the form of emails, photos,
areas such as bioinformatics, medical imaging, videos, monitoring devices, PDFs, audio, etc. are
sensor informatics, medical informatics and also being considered in the analysis
health informatics. applications.
2) Variety –Variety refers to heterogeneous depth overview of their target audience, draw
sources and the nature of data, both structured patterns and conclusions, and enhance their
decision-making. Media includes social media
2
and interactive platforms, like Google, IV. Health care and Big Data
Facebook, Twitter, YouTube, Instagram,
The Healthcare sector is booming at a faster rate
Cloud as a big data source - Today, companies and the necessity to manage patient care and
have moved ahead of traditional data sources by innovate medicines has increased synonymously.
shifting their data on the cloud. Cloud storage With the rise in such needs, newer technologies
accommodates structured and unstructured data are being adopted in the industry. One such major
and provides business with real-time information change is the use of Big Data and Analytics in the
and on-demand insights. As big data can be Healthcare sector.
stored and sourced on public or private clouds,
One of the characteristic that health care sector
via networks and servers, cloud makes for an
possesses is its data richness. With the
efficient and economical data source.
development in diagnostic and treatment, health
The web as a big data source - The public web care sector evolved so quickly in last few
constitutes big data that is widespread and easily decades. There are many sources in this sector
accessible. Data on the Web or ‘Internet’ is from where the data is generated. These data is
commonly available to individuals and undoubtedly in the form of Big Data. The data
companies alike. Moreover, web services such as came from many sources and categorized as
Wikipedia provide free and quick informational follows:
insights to everyone. 1) Web and social media data: Data captured
Iot as a big data source - Machine-generated from Facebook, Twitter, LinkedIn, blogs, and the
content or data created from IoT constitute a like. It can also include health plan websites,
valuable source of big data. This data is usually smartphone apps etc.
generated from the sensors that are connected to 2) Machine-to-machine (M2M) device generated
electronic devices. With IoT, data can now be data: readings from remote sensors, meters, and
processes, video games, meters, cameras, 3) Biometric data: Data may in form of retinal
household appliances, and the like. scans, x-ray images, finger prints, genetics,
handwriting, other medical images, blood
Databases as a big data source - Businesses pressure and other similar types of data .
today prefer to use an amalgamation of 4) Human-generated data: In the form of
traditional and modern databases to acquire unstructured and semi-structured data. Some of
relevant big data. Popular databases include a the examples are EMRs, Doctor’s notes and
variety of data sources, such as MS Access, paper documents.
DB2, Oracle, SQL, and Amazon Simple, among Genomic Data: data in the form of DNA
others. sequence.
3
V. Big Data Analytics sources to a data product useful for
organizations forms the core of Big Data
Big Data Analytics largely involves collecting
Analytics. Big Data analytics in Healthcare is
data from different sources, manipulate it in a
fundamentally a set of methodologies,
way that it becomes available to be consumed
procedures, frameworks and technologies
by analysts and finally deliver data products
which are used to transform raw data into
useful to the organization business.
meaningful as well as useful information. These
The process of converting large amounts of information are used to make decision making
unstructured raw data, retrieved from different effective
As per the figure 1, data produces from variety like extraction, cleaning, conformation,
of sources like hospitals, medical groups, payers transformation and loading are executed on data
or other data providers. These data first needed during this phase. Finally, some meaningful and
to be aggregated. Moreover, many processes useful information are generated which can be
4
used by variety of users and purpose as shown in costs and even eliminate the risk of chronic
figure 1. diseases.
5
8) Health Trend Analysis - By using different structured data is mostly heterogeneous. These
analytical approaches including data mining and may lead a huge problem at the time of
text mining techniques, health trend analysis and aggregation of these data. Natural language
comprehensive patient management is more easy processing and free-text software could solve
using Big Data Analytics. this problem at some extent but it is in its initial
stage.
9) Studying Drug Efficacy - Electronic health
record (EHR) data may also be used to study 3) Cloud Storage - The cloud storage can be used
drug efficacy. to upload data or having the whole system
designed in the cloud. Thus, the cloud will need
VII. Challenges and Solutions of Big to have sufficient space for the storage and
Data in Health Care sufficient speed for data upload at the same time.
1) Security - Since the big data contained The storage apart involving words
subject’s personal information and their health documentations, it should also able to store
history, it is important for the database to be graphic type such as X ray, CT or MRI. The
protected from hacking, cyber theft and phishing, system should also be able to generate graphic
where the stolen data can be sold for a huge sum. presentations from the available data so that
Before big data can be implemented, it is clinicians are able to visualize and understand
necessary to ensure that the administration, quickly and take prompt decision . the
privacy, security of the big data is well protected. advancement in cloud storage technology is
To address these security and privacy challenges, offering a potential solution to this problem
the big data analytics software solutions should through its added capacities of information
2) Data Aggregation - In health care sector, the routines, prioritise valuable data and train their
data is in unstructured form. These unstructured clinicians to recognise the value of relevant
6
5) Staying Up-to-Date - The dynamic nature of typically historical data or limited to government
healthcare data demands regular updations to payers.
keep it relevant. The time interval between each
8) Cost Incurred for Establishment of Big Data
update may vary from seconds to a couple of
Architecture - To have a benefit through Big
years for different datasets. It would be
Data analytics, it requires organization level
challenging to understand the volatility of big
management and analysis as well as a large-scale
data one is handling unless a consistent
investment.
monitoring process is in place.
6) Requirement of Expert Knowledge - Big Data VII. Technology Support for Big Data
systems require data scientists with specialized Analytics in Health Care
experience to support design, implementation,
and continued use. The McKinsey Global There are varieties of platforms and tools are
Institute estimates that there will be more than available for Big Data analytics in healthcare.
that mean 50–60% of data scientist positions 1) Cloud Storage - Cloud storage uses a network
may go vacant. Data scientists need highly of remote servers. These servers are hosted on
technical skill sets. They must possess soft skills the Internet to store, manage, and process data.
such as communication, collaboration, There are many vendors that provide cloud
leadership, creativity and more. storage. For example Google Cloud Storage is a
7) Protecting the Patient’s Privacy - One of the key part of storing and working with Big Data on
significant challenges in leveraging health care’s Google Cloud Platform. For Bigquery and
big data to its full extent is policies that protect Hadoop, using a Google Cloud Storage bucket is
the privacy of patient’s data. Many laws protect optional but recommended.
the patient’s data and not reveal the patient’s 2) Column oriented databases - Column-oriented
identity that makes the big data analytics databases basically stores data sets as segments
difficult. On the contrary, sometimes health care of columns of data rather than as rows of data. It
providers are themselves are reluctant to share allows huge data compression and very fast
data because of market competition. A physician query times.
many not want their competitors to know exactly
how many and which types of procedures they 3) NoSQL databases - In relational databases
performed and where. Also, the demographics of tabular relations are used while a NoSQL (Not
hospitals provide one hospital a financial only SQL) database provides a different method
advantage over another. Some of the datasets are for storage and retrieval of data. It focuses on
publicly available but these data sources are storage and retrieval of huge volumes of
7
semistructured, unstructured or even structured X - References
data.
1) Big data analytics -
4) Hadoop System - Hadoop is so far the most https://www.tutorialspoint.com/big_data_analyti
popular implementation of MapReduce cs/index.htm
methodology. It is an entirely open source
2) Bradley P. Implications of big data analytics
platform for handling Big Data.
on population health management. Big Data.
5)Hive - Hive is a runtime Hadoop support 2013;1(3):152–159.
architecture that leverages Structure Query
3) Big data evolution-
Language (SQL) with the Hadoop platform.
https://www.fingent.com/blog/5-ways-big-data-
6) PIG - PIG consists of a Perl-like language. is-changing-the-healthcare-industry
Instead of a SQL-like language, it allows for
4) big data and health care-
query execution over data stored on a Hadoop
https://practicalanalytics.co/2013/07/15/informat
cluster. Cassandra Cassandra is also a distributed
ics-or-analytics-understanding-healthcare-
database system. It is designated as a toplevel
provider-use-cases/
project modelled to handle big data distributed
across many utility servers. 5) Opportunities of big data in health care
https://www.villanovau.com/resources/bi/big-
IX. Conclusion
data-healthcare-opportunities/#.vnfrarz95kg
We may consider Big data as a latest evolution
6) Challenges of big data in health care-
in the field of decision support data management
http://healthcare-
systems. On the other side, the digitalization in
communications.imedpub.com/the-usefulness-
health care sector is in peak. As we discussed in
and-challenges-of-big-data-in-healthcare.pdf
the paper, there are several opportunities for Big
data in health care sector. Meanwhile, the 7)Advantages and challenges-
technological advancement is rapidly going on https://imarticus.org/healthcares-top-10-
towards the implementation of Big data challenges-in-big-data-analytics/
analytics. In near future, there will be widespread
8) Doug Laney, Application Delivery Strategies,
implementation of big data analytics across the
retrieved from http://blogs.gartner.com
health care organization and the healthcare
/douglaney/files/2012/01/ad949-3D-Data-
industry. The Big data solutions could definitely
Management-Controlling-Data-Volume-
save millions of life and improve patient
Velocity-andVariety.pdf on December 20, 2015
services.
9)Big data and health care-
https://catalyst.nejm.org/big-data-healthcare/
8
10) Big data characteristics-
https://www.guru99.com/what-is-big-data.html
16) https://www.ijstr.org/final-
print/mar2017/Improving-Healthcare-Using-Big-
Data-Analytics.pdf.