Application of Machine Learning in healthcare

1st Osama Yasser Elmihy 2nd Muhamed Hamdy Abdellah 3rd Mohamed Alaa Ahmed Refaey
Faculty of Engineering at Shobra Faculty of Engineering at Shobra Faculty of Engineering at Shobra
Benha University Benha University Benha University
Egypt Egypt Egypt
[email protected] [email protected] [email protected]

4th Mohamed Mazhar Mohamed Fadl 5th Mohamed Abdelmoneim Mahmoud

Faculty of Engineering at Shobra Faculty of Engineering at Shobra
Benha University Benha University
Egypt Egypt
[email protected] [email protected]

Abstract—There is a big effect on healthcare by Machine I. I NTRODUCTION

Learning (ML). ML considered as a subtype of Artificial In-
telligence (AI) technology. AI gives physicians more speed and The term ML is a branch of AI that refers to different
accuracy in work. AI can make it easy for countries to use techniques. It is concerned with development of algorithms
healthcare data to identify the trial sample which near to the that enable computers to have the ability to learn and help
ideal sample. ML techniques can assist in detecting indicators of to find solutions to problems. A ML system may identify or
an epidemic. ML collect data from social media reports or posts, recognize faces by studying a group of various pictures. [1]
videos and satellite data to make sure if the epidemic will become
out of control. Using ML for healthcare can give time for world Fields in which ML is used : engineering - medical - many
to focus on patient care instead of searching or entering data. of fields.
This report studies: Healthcare is one of the largest industries in the world that
A) Need for Machine Learning in healthcare. can benefit from technology. and it has always been a strong
B) Machine Learning efficiency in Healthcare. supporter of advanced technologies.
C) Associate features of Machine Learning for healthcare Benefits of ML in healthcare systems:
structure.
D) Pillars of Machine Learning for healthcare. • Improve the quality of automation,and providing the best
E) Significant applications of Machine Learning for health- care for patient, are used in case of emergency, and can
care. improve the quality of life for many of people in the
Index Terms—Machine Learning - Healthcare - Patient data world. item: reducing the expenses and time required for
medical tests.
• There are many of ML applications that can improve the
ARTIFICAL STORE
INTELLIGANCE DATA
effectiveness of clinical trials.
• Deep learning can analyze data that includes laboratory
tests, clinical notes, and medications at high speed with
possible accuracy. [3]

II. M ACHINE L EARNING IN H EALTHCARE

SATELLITE
SOCIAL MEDIA
DATA
REPORTS A. Need for Machine Learning in healthcare
The quality of healthcare services and the treatment with
Complex diseases are constantly improving. However, there
are still challenges, such: the dosage and duration of therapies
DEEP based on individual characteristics of the patient. but, in recent
DATA LEARNING
ANALYSIS years, ML has been successfully integrated into pediatric care
to provide the best care for children. ML has helped many
hospitals and health systems deal with unique challenges. It
is worth mentioning that ML technology is one of the most
Fig. 1. Machine Learning Applications
exciting areas of AI, and many companies are taking advan-
tage of ML technology for their purposes. ML is becoming
increasingly popular. It uses algorithms to facilitate data driven
learning. Healthcare is constantly changing due to the constant
development of new technology and ideas. ML could assist have plagued traditional organizations for decades, and ML
medical professionals in some of these new scenarios [4]. techniques are already part of the solution. This is because of
vast databases and intelligent search algorithms. [5]
B. Machine Learning efficiency in Healthcare
D. Pillars of Machine Learning for healthcare
The advancement of this technology has hastened the tran-
sition to information-based healthcare delivery and adminis- The concept of ML and its versatile capabilities have been
tration. Another advantage of employing ML techniques in reported to serve the healthcare domain in various ways.
healthcare is that it decreases the likelihood of human error The outbreak prediction capability, medical imaging diagnosis,
by removing some human participation This is especially true behavioral modifications, records of patient data, etc..., are
for process automation activities, as tiresome repetitive work some
is where humans make the most mistakes. Clinical decision ML is the process of teaching computers to recognize
support technologies aid in analyzing huge amounts of data patterns by supplying them with data and an algorithm to work
to detect a condition, decide on the next stage of treatment, with. ML plays a significant part in diagnosing the patient’s
identify any potential difficulties, and improve overall patient ailment, monitoring his health, and recommending the required
care efficiency. So, this technique is powerful that assists steps to prevent it. It can range from minor illnesses to severe
physicians in doing their duties more efficiently and promptly diseases , such as cancer, which is difficult to detect early.
and reduces the likelihood of making an incorrect diagnosis or This technology can analyze vast amounts of data gath-
prescribing ineffective treatment and its growing popularity in ered from real-time social media feeds, satellites, website
recent years. The cause for this is the increased acceptance of information, and institutional records. Networks can assist
electronic health records and one of the most important of it in making sense of this data and forecasting anything from
is medical photographs. Medical images, such as X-rays, were malaria outbreaks to serious infectious diseases.
analogue for the longest time. ML enables machines to learn concepts, interpret data, and
provide desired outputs independently. And its models learn
C. Associate features of Machine Learning for healthcare to interpret data through conditions and clauses using various
structure learning approaches such as unsupervised, supervised, and
more. As a result, they are suitable for producing prescriptive
and predictive insights.
Furthermore, ML aids in optimizing patient involvement
and recovery by sending out timely alerts and notifications
to patients regarding their visits, report collection, and other
activities. This technology can analyze vast amounts of data
gathered from real-time social media feeds, satellites, website
information, and institutional records. Networks can assist in
making sense of this data.
Daily, healthcare workers spend their time on duplicate
chores like record administration and upkeep, and claims
processing, which stops them from providing the necessary
treatment. The use of ML models may result in automation
Fig. 2. Associate features of Machine Learning for healthcare structure and the elimination of human intervention in locations. ML
has become an inseparable part of industries, bringing signifi-
There are variety of smart and caring features associated cant advancements that benefit entrepreneurs by making their
with the ML technology for its full services in healthcare businesses more profitable and customer-centric.
perspectives. It includes support from the various intelligent ML helps healthcare professionals with various tools and
and digital tools like AI and cloud data performances be- techniques that significantly impact health outcomes. And its
longing to healthcare services. In addition to that, its cost software, equipped with predictive algorithms, can provide
is reasonable. ML is helping in generating of electronic immediate benefits to disciplines with large datasets. [2]
medical records further Which assists the healthcare domain.
The smart prepared report, digital notes, records maintaining, E. Significant applications of Machine Learning for health-
etc., are several other impactful zones. Healthcare institutions care
are using ML systems to monitor and anticipate potential 1) These technologies use algorithms that can extract infor-
epidemic outbreaks in various parts of the world. This digital mation from clinical charts faster and more accurately
system can forecast disease outbreaks by receiving data from than manual review processes, and they become more
satellites, real-time updates on other crucial information from brilliant as more documents are processed.
the web. It is a boon for third-world countries that lack 2) ML allows plans and providers to identify hidden risk
adequate healthcare facilities. It solves complex appointment factors and healthcare gaps and improve risk score
process, and a lack of access to the correct. Similar challenges accuracy.
 3) ML in healthcare delivers algorithms with self-learning whole world. The most challenging matter in ML development
neural networks that can improve treatment quality by is obtaining patients’ data because it is safeguarded by tough
assessing external data on a patient’s condition, X-rays, privacy and security restrictions. Furthermore, data format and
CT scans, and numerous tests and screenings. quality issues need a big amount of effort to clean and make
4) Algorithms outperform radiologists in detecting malig- it ready for ML analysis.
nant tumors. [6]
IV. FUTURE VISION
III. DISCUSSION smart medical devices grow common. as a result,
A. THE ROLE OF Machine Learning IN SOCIETY technology-enabled health care is becoming a reality. This
Today, the whole world is affected by ML. and as it is in technology is responsible for many definite data points, accu-
progress every day it impacts significantly global applications rately predicting risks and outcomes, and performing various
from car self-driving to medical research and health care. All other duties. in the future this can be improved to make
these applications depend on the progress of ML algorithms. a personalized prescription for people having unique sets
Therefore, developers and researchers bear a big deal of of particular needs. ML helps with the nowadays issue and
responsibility to keep up with world needs. ML is indispens- anticipates future difficulties as well as outbreaks across the
able for lap diagnosis, intelligent clinical data analysis, and world. the experts must collect a large amount of data from
computer-aided medical procedures such as organ recognition. websites, real-time social media, and other today’s world. It
ML is different from traditional explicitly programmed assists in verifying these data and predicting small things to
models in which the ML systems are trained. Some ML for dilemmas like serious infectious diseases.
health care and research will process personal data as part of V. CONCLUSION
the model’s function. This technology is ready to improve the
ML can be a powerful tool in the hands of a health
caring process and patients’ lives starting from diagnostics.
practitioner, scientist, or researcher. Every day, it appears that
The increased reliance on ML over a wide range results in
there is a breakthrough in ML. With each new breakthrough,
big amounts of data that ML automation needs for effective
a new ML application emerges that may solve a real problem.
processing and because the wealth of information can help
ML standards are supporting medical doctors and surgeons in
professionals to make sound decisions.
saving lives, and detecting illnesses and concerns even earlier
Deep learning algorithms provide an accurate and efficient
than they occur.
diagnostic suggestion. as a result they reduce care costs, reduce
Corporations now can accelerate their testing and commen-
the administrative load of healthcare professionals, allow them
tary by means of the usage of digital medical trials and patterns
to focus more on other complex diagnostics, and prevent
of identity. Companies recognize that in order to improve
delays in reporting urgent cases.
general fitness, they must address the whole person, including
B. HOW TO EFFICIENTLY USE Machine Learning IN lifestyle and environment. ML models can identify people at
HEALTH CARE higher risk of developing preventable chronic diseases like
All the professionals need to understand ML chances and heart disease and diabetes.
they will be able to improve clinical outcomes and the hospital R EFERENCES
experience. ML automates computer systems by recognizing
[1] M. A. Ahmad, C. Eckert, and A. Teredesai, “Interpretable machine
patterns and converting unstructured data to structured ones. learning in healthcare,” in Proceedings of the 2018 ACM International
Practitioners benefit from that to make their decisions in a Conference on Bioinformatics, Computational Biology, and Health
short time. Informatics, ser. BCB ’18. New York, NY, USA: Association
for Computing Machinery, 2018, p. 559–560. [Online]. Available:
ML plays an important role in medical decision support, https://doi.org/10.1145/3233547.3233667
documentation of diseases in a short time, and providing [2] S.-Y. Hsu, Y. Ho, P.-Y. Chang, C. Su, and C.-Y. Lee, “A 48.6-to-105.2
treatment plans for optimal outcomes. Some deep learning µw machine learning assisted cardiac sensor soc for mobile healthcare
applications,” Solid-State Circuits, IEEE Journal of, vol. 49, pp. 801–
techniques such as image recognition and object detection 811, 04 2014.
are used in magnetic resonance and computed tomography [3] S. Siddique and J. C. L. Chow, “Machine learning in healthcare
to detect and predict diseases. ML developments have yielded communication,” Encyclopedia, vol. 1, no. 1, pp. 220–239, 2021.
[Online]. Available: https://www.mdpi.com/2673-8392/1/1/21
better dealing with patients’ data, the development of novel [4] S. Vyas, M. Gupta, and R. K. Yadav, “Converging blockchain and
medical procedures, and better treatment of chronic diseases. machine learning for healthcare,” 2019 Amity International Conference
The amount of available data has increased rapidly in recent on Artificial Intelligence (AICAI), pp. 709–711, 2019.
[5] A. Callahan and N. H. Shah, “Chapter 19 - machine learning
years, however, ML systems are based on enormous datasets, in healthcare,” in Key Advances in Clinical Informatics,
and as it advances it will be more integrated with these A. Sheikh, K. M. Cresswell, A. Wright, and D. W. Bates,
data. So hospitals must have good, dependable, and well- Eds. Academic Press, 2017, pp. 279–291. [Online]. Available:
https://www.sciencedirect.com/science/article/pii/B9780128095232000194
timed data for AI and ML to not lose their effectiveness. The [6] R. Kumar, A. R. Verma, M. K. Panda, and P. Kumar, “Hrv signal feature
quality of automation and intelligent decision-making should estimation and classification for healthcare system based on machine
be improved in hospitals and public healthcare organizations learning,” Communications in Computer and Information Science, 2020.
to improve the life quality of billions of people around the