Machine Learning in Medicine A Primer

International Journal of Trend in Scientific Research and Development (IJTSRD)
Volume: 3 | Issue: 2 | Jan-Feb 2019 Available Online: www.ijtsrd.com e-ISSN: 2456 - 6470
@ IJTSRD | Unique Reference Paper ID - IJTSRD20255 | Volume – 3 | Issue – 2 | Jan-Feb 2019 Page: 98
Machine Learning in Medicine: A Primer
Matthew N. O Sadiku1, Sarhan M. Musa1, Adedamola Omotoso2
1Professor, 2Student
Roy G. Perry College of Engineering, Prairie View A&M University, Prairie View
ABSTRACT
Machine learning is the field that focuses on how computerslearn fromdata.Today,machinelearningis playingan integral role
in the medical industry. This is due to its ability to process huge datasets beyond the scope of human capability, and then
convert the data analyzed into clinical insights that aid physicians in providing care. Machine learning is a powerful,relatively
easy to implement tool with numerous possibilities to enhance medical practice. The applications of machine learning in
medicine are advancing medicine into a new realm. Therefore, educating the next generation of medical professionals with
machine learning is essential. This paper provides a brief introduction to applying machine learning in medicine.
KEYWORDS: machine learning, computer algorithms, artificial intelligence, medicine, healthcare
INTRODUCTION
In today’s information era, an organization should deduct
meaningful information from their data obtained in orderto
be able to compete. Like other fields, medicine is going
through data revolution. With the emergence big data, the
explosion in data available for analysis is as evident in
healthcare as anywhere else. Clinical medicine has always
required doctors to handle enormous amounts of data.
Diagnostic and imaging techniques generate such an
incredible amount of data. Machine Learning techniques
emerged as an objective tool to assist practitioners to
diagnose certain conditions and make clinical decisions [1].
Most computer-based algorithms in medicine are “expert
systems.” Expert systems work the way an ideal medical
student would: they take general medical principles and
apply them to patients. Machine learning handles problems
as a doctor progressingthrough residencymight:bylearning
rules from data. It is a branch in computer science that
employs knowledge from artiﬁcialintelligence,optimization,
and statistics to develop algorithms. It focuses on how
computers learn from data. Such an algorithmic method
allows data to speak for themselves.
Although machine learning (ML) is widely used in other
industries, such as retail and banking, it is not routine in
medicine because of the complexity and limited availability
of data. In medicine, the bottom line is to use machine
learning to augment patient care, save more lives, improve
more care, while saving money at the same time. ML can
automate the manual processes carried out bypractitioners,
which are usually time-consuming and subjective. Thus,
using ML can save time for practitioners and provide
unbiased, repeatable results.
MACHINE LEARNING
Machine learning (ML) is the discipline that gives computers
the ability to learn without being explicitly programmed. The
term “machine learning” (ML)wasinitiallycoinedin1959 by
Arthur Samuel, a computer scientist. Machine learning (or
statistical learning) is part of artificial intelligence. As
illustrated in Figure 1, ML may also be regarded as a sub-
disciplines of data science [2]. It assists computers in
estimating future events and modelling based on
experiences gained from previous information. Machine
learning (ML) focusesonhowcomputers“learn”fromdata.It
allows computers to learn from past examples and to detect
hard-to-discern patterns from largedatasets. An exampleof
how a machine learner is trained is shown in Figure 2 [3]. It
describes a class of algorithms which learn model
parameters from a set of training data with the purpose of
accurately predicting outcomes for previously unseen data.
ML is a marriage between statistics and computer science
[4,5].
There are two types of learning: supervised learning and
unsupervised learning. Supervised learning focuses on
classification and prediction. It involves buildingastatistical
model for predicting or estimating an outcome basedonone
or more inputs. It is often used to estimate risk. Supervised
ML is where algorithms are given training data. Learning
from data is used when there is no theoretical or prior
knowledge solution, but data is available to construct an
empirical solution. Supervised ML is increasinglybeingused
in medicine such as in cardiac electrophysiology. In
unsupervised learning, we areinterested in findingnaturally
occurring patterns within the data. Unlike supervised
learning, there is no predicted outcome. Unsupervised
learning looks for internal structure in the data [6].
Unsupervised learning algorithms are common in neural
network models. A common application of such a process is
to explore interrelationships between genetics,
biochemistry, histology, and disease states.
The most commonly used ML algorithms include neural
networks, support vector machines, and decision trees.
Other ML techniques include classiﬁcation, regression, and
clustering. There are some software packages that do some
form of machine learning [7].
There is significant interest in the use of ML in medicine. ML
techniques can ‘learn’ from the vast amount of medical data
and assist clinical decision making. They are often suitable
for detecting complex patterns in large and noisy data sets.
For a ML system to be useful in solving medical diagnostic
tasks, the following features are desired: good performance,
the ability to appropriately deal with missing data, the
transparency of diagnostic knowledge, the ability to explain
decisions, and the ability of the algorithm to reduce the

International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
number of tests necessary to obtain reliable diagnosis [8].
However, success in using ML is not always guaranteed. Like
any technique, an appreciation of the limitations of ML
algorithms is crucial.
APPLICATIONS OF ML IN MEDICINE
Machine learning techniques have been currently applied in
the analysis of data in various fields including medicine,
finance, business, education, advertising, cybersecurity, and
energy applications. They were first applied to medicine
with the use of electronic health records. They emerged in
the medical sciences as clinical decision-support techniques
to improve sensitivity, disease detection, and monitoring.
There are almost limitless opportunities for machine
learning in medicine. We can considersomepopular areas in
medicine that have benefited or might benefit frommachine
learning approaches. Both supervised and unsupervisedcan
be applied to clinical data sets for the purpose of developing
robust risk models and redefining patient classes. Few
applications of machine learning include [9, 10]:
Oncology: Almost all works in this field apply machine
learning techniques, which perform deep statistical
analysis of a set of clinical cases supported by gene
expression data. Researchers are using deep learningto
train algorithms to recognize cancerous tissue at a level
comparable to trained physicians.Googlehasdeveloped
a machine learning algorithm to help identify cancerous
tumors on mammograms. Stanford is using a deep
learning algorithm to identify skin cancer. Lately
machine learning has been applied to cancer prognosis
and prediction [3].
Pathology: Pathology is the medical specialty that is
concerned with the diagnosis of disease. Machine vision
and other machine learning technologies can enhance
the efforts traditionally left only to pathologists with
microscopes.
Genomic Medicine: One of the goals of genomic
medicine is to determine how variations in the DNA of
individuals can affect the risk of different diseases.
Machine learning can help to model the relationship
between DNA and the quantities of key molecules in the
cell [11].
Predicting ChronicDisease: Machine learningcan help
hospital systems identify patients with chronic disease,
predict the likelihood that patients will develop chronic
disease, and present patient-specific prevention
interventions. For example, ML approaches have
significant potential in prediction of acute exacerbations
in asthma patients [12].
Personalized Medicine: This is individualization,
recognition of the micro-variables within a patient that
may cause them to be different from their peers. ML for
personalized medicine is a growing area. It involves the
ability to draw on large data sets and predictive models
allows for clinicians to more confidently diagnose,
predict and treat their patients.
Reduce Readmissions: Machine learning can reduce
readmissions. Clinicians can receive dailyguidanceasto
which patients are most likely tobereadmittedand how
they might be able to reduce that risk.
These advances would have been unimaginable without
machine learning. Other areas include surgery, radiology,
emergency medicine, traditional medicine, mental health,
psychiatry, ophthalmology, and optometry. ML provides
methods, techniques, and toolsthatcanhelpsolvediagnostic
and prognostic problems in various medical domains.
Machine learning techniques have been successfully applied
to different branches of medicine as a tool to help diagnose
diseases. ML algorithms have immense potentialto enhance
diagnostic andintervention research insmoking,depression,
asthma, and chronic obstructive pulmonarydisease(COPD).
ML is also being used for data analysis and interpretation of
continuous data used in the Intensive Care Unit [13]. With
time, the healthcare environment will become more and
more reliant on computer technology and ML capabilities
will reach into all aspects of medicine.
BENEFITS AND CHALLENGES
Machine accuracy will soon exceed that of humans. Machine
learning can be trained to look at images, identify
abnormalities, and point to areas that need attention. It can
be helpful in alerting a physician or a patient’s family when
things begin to go wrong. It can lead to more accurate
diagnostic algorithms and individualize patienttreatment. It
has the potential to positively change the role of physicians
in patient care.
While ML has the potential to significantly help medical
practice, some negative consequences that may arise from
using ML decision support in medicine. Critical voices have
emerged warning of potential problemssurroundingtheuse
of ML [14]. Ethical and regulatory concerns about health
data have expressed for long. ML algorithms use data that
are subject to privacy protections, requiring that their
developers pay close attention to ethical and regulatory
restrictions. Data used to train algorithms should have the
necessary use authorizations, but determining which data
uses are permitted for a given purpose is not an easy [15].
People are understandably hesitant when they learn that
healthcare professionals are being replaced by algorithms.
Machine learning algorithms often require thousands of
observations to reach acceptable performance levels. How
doctors handle these challenges will create winners and
losers in medicine.
CONCLUSION
Machine learning should no longer be regarded as a
futuristic concept but a real-world tool that can be deployed
today. Machine learning is critical for anyone practicing
medicine in the 21st century. Today, ML techniques are
suited for analyzing medical data and serving as tools for
medical diagnosis in a variety of medical domains. They will
become indispensable tools for clinicians who truly seek to
understand their patients.
If technology is to improve medicine in the future, the data
provided to doctors needs to be enhanced by the power of
analytics and machine learning. This will help them make
better decisions about patient diagnoses and treatment
option. Machine learning represents the next wave in
advancing modern medicine. The next generation of medical
professionals should be equipped with the right ML
techniques that will enable them to become part of this
emerging revolutionary technology [16]. For more
information on ML, see the book in [17] and other books
available at Amazon.com

International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
REFERENCES
[1] A. Consejo, T. Melcer, and J. J. Rozema,“Introduction to
machine learning for ophthalmologists, seminars in
ophthalmology,” November 2018.
[2] V. Bolon-Caned et al., “Machine learning for medical
applications,” Proceedings of European Symposium on
Artificial Neural Networks, Computational Intelligence
and Machine Learning, Belgium, April 2016, pp. 225-
234.
[3] J. A. Cruz and D. S. Wishart, “Applications of machine
learning in cancer prediction and prognosis,” Cancer
Informatics, vol. 2, 2006, pp. 59–78.
[4] Z. Obermeyer and E. J. Emanuel, “Predicting thefuture
— Big data, machine learning, and clinical medicine,”
New England Journal of Medicine, vol. 375, no. 13,
September 2016, pp. 1216–1219.
[5] M. N. O. Sadiku, S. M. Musa, and O. S. Musa, “Machine
learning, ” International Research Journal of Advanced
Engineering and Science, vol. 2, no. 4, 2017, pp. 79-81.
[6] R. C. Deo, “Machine learning in medicine,” Circulation,
vol. 132, no. 20, November 2015, pp. 1920–1930.
[7] A. K. Waljee and P. D. R. Higgins, “Machine learning in
medicine: A primer for physicians,” The American
Journal of Gastroenterology, vol. 105, June 2010, pp.
1224–1226.
[8] I. Kononenko, “Machine learningfor medicaldiagnosis:
History, state of the art and perspective,” Artificial
Intelligence in Medicine, vol. 23, no. 1, August 2001, pp.
89-109.
[9] K. Sennaar, ”Machine learningformedicaldiagnostics–
4 current applications,” December 12, 2018,
https://emerj.com/ai-sector-overviews/machine-
learning-medical-diagnostics-4-current-applications/
[10] G. S. Handelman, “eDoctor: Machine learning and the
future of medicine,” Journal of Internal Medicine, vol.
284, no. 6, December 2018, pp. 603-619.
[11] M. K. K. Leung et al., “Machine learning in genomic
medicine: A review of computational problems and
data set,” Proceedings of the IEEE, vol. 104, no. 1,
January 2016, pp. 176-197.
[12] J. Finkelstein1 and I. C. Jeong, “Machine learning
approaches to personalize early prediction of asthma
exacerbations,” Annals of the New York Academy of
Sciences, vol. 1387, 2017, pp. 153–165.
[13] G. D. Magoulas and A. Prentza, “Machine learning in
medical applications,” chapter in G. Paliouras, V.
Karkaletsis, and C.D. Spyropoulos (eds.): Advanced
Course on Artificial Intelligence, 1999, Springer, pp.
300-307.
[14] L. McDonald et al., “Unintended consequences of
machine learning in medicine? “ F1000Research, vol. 6,
2017.
[15] E. Vayena, A. Blasimme, and I. G. Cohen, “Machine
learning in medicine: Addressing ethical challenges,”
PLoS Medicine, vol. 15, no. 11, 2018.
[16] V. B. Kolachalama and P. S. Garg. “Machine learning and
medical education,” npj Digital Medicine, vol. 54,
September 2018.
[17] T. J. Cleophas and A. H. Zwinderman, Machine
Learning in Medicine, Part 2. Springer, 2018.
Figure 1 (a) Data science, (b) Machine learning [2].
Figure 2 an example of how a machine learner is trained [3].

Machine Learning in Medicine A Primer

More Related Content

What's hot (20)

Similar to Machine Learning in Medicine A Primer (20)

More from ijtsrd (20)

Recently uploaded (20)

Machine Learning in Medicine A Primer