Machine Learning in Healthcare
Machine Learning in Healthcare
Abstract: This paper discusses machine role in different fields involved with our
learning and its role in general, and also lives, such as, video surveillance,
discusses its growth and current computing, robotics, manufacturing
advancement in the field of healthcare. industries, social media, etc. (1) In this
Lastly, it discusses some of the article, we will discuss the involvement of
associated problems of machine machine learning in another very important
learning. Machine learning is the field, and that is, healthcare.
application of Artificial Intelligence that
allows the system to work by itself
without being programmed. Machine
Role of Machine
Learning contributes in technological Learning
related aspects of many fields like
computing, robotics, sports analysis,
healthcare, etc. Machine learning has As shortly mentioned in the previous
some associated problems as well, paragraph, Machine learning plays a role in
however they are not significant developing computer programs that can
compared to its good sides. access data to use them to learn for
themselves. Machine learning is used in
Keywords: algorithm, programming, AI. various aspects of the current world.
Industries are developing stronger machine
learning models that can analyze bigger
Introduction and more complex data with a higher speed
delivery and more accurate results.
Machine learning means the application of Systems using machine learning enables
Artificial Intelligence that enables the the government officials to use data to
system to automatically work without predict likely future situations and adapt to
through programming. Machine learning that rapidly. Machine learning can help the
works in developing computer programs government to improve cyber security and
that can access data to use them to learn for cyber intelligence as well. It can help the
themselves. In machine learning, the counterterrorism efforts, optimize
process begins with observations or data, operational preparedness, logistic
for instance, direct experience, or management, predictive maintenance and
instruction in order to find out patterns in can reduce failure rates. Machine learning
data to make better decisions based on the plays a role in the marketing and sales
given examples. The main goal is to enable industry as well. Many companies have
the computers to learn automatically implemented artificial intelligence and
without being controlled by humans. machine learning to increase customer
Machine learning has played a significant contentment by over 10%. As believed by
57% of the enterprise executives, most thousands of variables. The development
important growth benefits of AI and of machine learning is allowing the
machine learning will improve customer optimization of algorithms in real time. (3)
experience and support. E-commerce and
social media sites also use machine
learning. These sites use machine learning
Current Advancements
to analyze customers’ buying and search
There is a nice article written by Mike
history and recommend items based on
Thomas reflecting the current advantages
that. Transportation industry uses machine
machine learning made in healthcare and
learning as well. It's always important to
how it is being used. (4) The author names
maintain efficiency and accuracy in this
several companies and describes how they
industry. Therefore the ability to predict
are using machine learning in healthcare.
and mitigate potential problems are also
For example, Quotient Health, a company
important. Machine learning’s data
located in Denver, Colorado. Thomas
analysis and modeling functions links very
writes, “With the help of machine learning,
well with businesses within the delivery,
Quotient Health developed software that
public transportation, and freight
aims to “reduce the cost of supporting
transportation sectors. Machine learning
EMR [electronic medical records]
uses algorithms to find factors that impact
systems” by optimizing and standardizing
a supply chain’s success both in positive
the way those systems are designed. The
and negative ways which makes machine
ultimate goal is improved care at a lower
learning a critical component within supply
cost.” Then Thomas mentions another
chain management. (2)
company that uses machine learning.
According to what he stated, the company
Growth of Machine called KenSci uses machine learning to
predict illness and treatment in order to
Learning in Healthcare help the doctors and payers intervene
faster, predict population risk by
Use of Algorithms is not new in
identifying patterns and surfacing high risk
healthcare. The first algorithms were
areas, etc. After that, Thomas talks about
digital flow charts that depended on a
another company called Ciox Health which
small number of data points and binary
is located in Alpharetta, Georgia. About
choices provided by the user. In healthcare,
this company the author of the article
these early algorithms worked as physician
writes “Ciox Health uses machine learning
facilitators which are a series of rules-
to enhance "health information
based decision trees and checklists
management and exchange of health
designed for guiding and standardizing
information," with the goal of modernizing
care. However, over the past decade, the
workflows, facilitating access to clinical
growth in computing power has made it
data and improving the accuracy and flow
possible to make predictions based on
of health information.”
Associated Problems
Conclusion
Like every other good thing that benefits
us in many ways, machine learning has
Machine learning has brought
some associated problems or limitations as
revolutionary changes in the world of
well. Matthew Stewart, a PhD researcher,
technology. Various fields are using it to
wrote an article describing the limitations
further the current scientific and
of machine learning. In his article, the first
technological advancements in order to
limitation he mentioned is ethics. As
make people’s lives easier. If the problems
Stewart said, in machine learning we trust
associated with machine learning can be
algorithms and data more than our own
reduced to the lowest then it can become
judgement, and according to him, it has
more useful for us.
pros and cons. As for pros, then we benefit
from it, and as for cons, Stewart says that
we are replacing someone’s job with
algorithms, which comes with ethical [1] W. I. R. E. D. Insider, “From
ramifications. He also raises the question
Diagnosis to Holistic Patient Care,
that who should we blame if something
goes wrong. He writes, “The most Machine Learning is Transforming
commonly discussed case currently is self-
driving cars — how do we choose how the Healthcare,” Wired, 04-Dec-2019.
vehicle should react in the event of a fatal
collision? In the future will we have to [Online]. Available:
select which ethical framework we want
https://www.wired.com/wiredinsider
our self-driving car to follow when we are
purchasing the vehicle?” Next the author /2019/10/from-diagnosis-to-holistic-
discusses the second limitation which is
Deterministic Problems. As someone patient-care-machine-learning-is-
whose field of expertise is environmental
science, the author has to deal with this transforming-healthcare/. [Accessed:
kind of problem. As he stated, predicting
28-Jan-2021].
weather using machine learning is
problematic. He wrote, “Using a neural
network with a thousand inputs to [2] P. D. R. Matthew Stewart, “The
determine whether it will rain tomorrow in
Boston is possible. However, utilizing a Limitations of Machine Learning,”
neural network misses the entire physics of
Medium, 29-Jul-2020. [Online].
the weather system.” (5)
Available:
https://towardsdatascience.com/the- how-will-it-impact-business/.
2021].
Available:
https://builtin.com/artificial-
intelligence/machine-learning-
06-May-2020. .
[Online]. Available:
https://www.quantilus.com/why-is-
machine-learning-important-and-