Machine Learning and Deep Learning in Medical Data

Analytics and Healthcare Applications 1st Edition

Visit the link below to download the full version of this book:

https://medidownload.com/product/machine-learning-and-deep-learning-in-medical-d
ata-analytics-and-healthcare-applications-1st-edition/

Click Download Now

First edition published 2022
by CRC Press
6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742
and by CRC Press
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
CRC Press is an imprint of Taylor & Francis Group, LLC
© 2022 selection and editorial matter, Om Prakash Jena, Bharat Bhushan, Utku Kose; individual chapters,
the contributors
Reasonable efforts have been made to publish reliable data and information, but the author and publisher
cannot assume responsibility for the validity of all materials or the consequences of their use. The authors
and publishers have attempted to trace the copyright holders of all material reproduced in this publication and
apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright
material has not been acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or here-
after invented, including photocopying, microfilming, and recording, or in any information storage or
retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, access www.copyright.com
or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923,
978-750-8400. For works that are not available on CCC please contact [email protected]
Trademark notice: Product or corporate names may be trademarks or registered trademarks and are
used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Names: Jena, Om Prakash, editor. | Bhushan, Bharat, editor. | Kose, Utku, 1985- editor.
Title: Machine learning and deep learning in medical data analytics and
healthcare applications / edited by Om Prakash Jena, Bharat Bhushan, Utku Kose.
Description: First edition. | Boca Raton : CRC Press, 2022. | Series: Biomedical and robotics
healthcare | Includes bibliographical references and index. | Summary: “The book incorporates
the many facets of computational intelligence, such as machine learning and deep learning, to
provide groundbreaking developments in healthcare applications. It discusses theory, analytical
methods, numerical simulation, scientific techniques, analytical outcomes, and computational
structuring”-- Provided by publisher.
Identifiers: LCCN 2021043580 (print) | LCCN 2021043581 (ebook) | ISBN 9781032126876 (hardback) |
ISBN 9781032127644 (paperback) | ISBN 9781003226147 (ebook)
Subjects: LCSH: Medical technology. | Medical care--Technological innovations. |
Robotics in medicine.
Classification: LCC R855.3 .M33 2022 (print) | LCC R855.3 (ebook) |
DDC 610.285--dc23/eng/20211103
LC record available at https://lccn.loc.gov/2021043580
LC ebook record available at https://lccn.loc.gov/2021043581

ISBN: 978-1-032-12687-6 (hbk)

ISBN: 978-1-032-12764-4 (pbk)
ISBN: 978-1-003-22614-7 (ebk)

DOI: 10.1201/9781003226147

Typeset in Times LT Std

by KnowledgeWorks Global Ltd.
Contents
Preface......................................................................................................................vii
Editors........................................................................................................................ix
Contributors...............................................................................................................xi

Chapter 1 Common Data Interface for Sustainable Healthcare System................1

C. B. Abhilash, K. T. Deepak, Rajendra Hegadi,
and Kavi Mahesh

Chapter 2 Brain–Computer Interface: Review, Applications

and Challenges.................................................................................... 23
Prashant Sengar and Shawli Bardhan

Chapter 3 Three-Dimensional Reconstruction and Digital Printing

of Medical Objects in Purview of Clinical Applications.................... 39
Sushitha Susan Joseph and Aju D

Chapter 4 Medical Text and Image Processing: Applications, Methods,

Issues, and Challenges......................................................................... 65
Behzad Soleimani Neysiani and Hassan Homayoun

Chapter 5 Usage of ML Techniques for ASD Detection: A Comparative

Analysis of Various Classifiers........................................................... 91
Ashima Sindhu Mohanty, Priyadarsan Parida,
and Krishna Chandra Patra

Chapter 6 A Framework for Selection of Machine Learning

Algorithms Based on Performance Metrices and Akaike
Information Criteria in Healthcare, Telecommunication,
and Marketing Sector........................................................................ 113
A. K. Hamisu and K. Jasleen

Chapter 7 Hybrid Marine Predator Algorithm with Simulated

Annealing for Feature Selection....................................................... 131
Utkarsh Mahadeo Khaire, R. Dhanalakshmi,
and K. Balakrishnan

v
vi Contents

Chapter 8 Survey of Deep Learning Methods in Image Recognition

and Analysis of Intrauterine Residues............................................... 151
Bhawna Swarnkar, Nilay Khare,
and Manasi Gyanchandani

Chapter 9 A Comprehensive Survey on Breast Cancer Thermography

Classification Using Deep Neural Network...................................... 169
Amira Hassan Abed, Essam M. Shaaban,
Om Prakash Jena, and Ahmed A. Elngar

Chapter 10 Deep Learning Frameworks for Prediction,

Classification and Diagnosis of Alzheimer’s Disease....................... 183
Nitin Singh Rajput, Mithun Singh Rajput,
and Purnima Dey Sarkar

Chapter 11 Machine Learning Algorithms and COVID-19: A Step for

Predicting Future Pandemics with a Systematic Overview.............. 203
Madhumita Pal, Ruchi Tiwari, Kuldeep Dhama,
Smita Parija, Om Prakash Jena,
and Ranjan K. Mohapatra

Chapter 12 TRNetCoV: Transferred Learning-based ResNet Model

for COVID-19 Detection Using Chest X-ray Images........................ 219
G. V. Eswara Rao and B. Rajitha

Chapter 13 The Influence of COVID-19 on Air Pollution

and Human Health............................................................................ 235
L. Bouhachlaf, J. Mabrouki, and S. El Hajjaji

Chapter 14 Smart COVID-19 GeoStrategies using Spatial

Network Voronoï Diagrams.............................................................. 245
A. Mabrouk and A. Boulmakoul

Chapter 15 Healthcare Providers Recommender System Based

on Collaborative Filtering Techniques.............................................. 261
Abdelaaziz Hessane, Ahmed El Youssefi,
Yousef Farhaoui, Badraddine Aghoutane,
Noureddine Ait Ali, and Ayasha Malik
Index....................................................................................................................... 275
Preface
The most astonishing difference between computers and humans lies in the fact that
computers need to be programmed in order to respond to any event whereas humans
learn from their past experience. However, with the advent of machine learning (ML)
and deep learning (DL), it is possible for computers to learn from their experiences.
Recent advances in ML/DL algorithms are impervious to large-scale technological
disruptions and have transformed numerous industries such as governance, trans-
portation, manufacturing, and healthcare. These techniques have shown tremendous
results in varied healthcare-related tasks such as brain tumor segmentation, medi-
cal image reconstruction, lung nodule detection, classification of lung diseases, and
medical image recognition. Furthermore, the exponentially growing volume of bio-
medical big data generated due to health data collection through digital health wear-
ables, genomic sequencing, and electronic health records (EHRs) is another matter
of concern. ML/DL schemes have a proven ability to extract actionable knowledge
from these large health datasets. ML models can also contribute toward improving
the quality of care, enhancing patient safety, and mitigating the overall healthcare
costs. Extraction of appropriate data would be extremely beneficial in resolving seri-
ous medical conditions to a significant extent. ML/DL approaches can be used to
extract certain attributes, and the trained model can be used to make proper diag-
noses and prognoses from available medical data and photographs. These can also
ease the identification of high-risk patients, early detection of lung cancer, detec-
tion of abusive and fraudulent health insurance claims, and diagnosis of respiratory
ailments from chest X-rays. ML and big data strategies are used to build predic-
tive diagnostic systems on collected data. However, designing and implementing an
effective diagnostic system remains a difficult task due to a variety of issues such as
stability, accessibility, scalability, safety, development standards, and technologies.
This book covers the fundamentals of ML and DL in the healthcare domain where
these models are used to train the system and implicitly extract positive solutions.
The main aim of this book is to highlight the role of ML/DL algorithms in improved
healthcare diagnostic systems, processing EHRs, medical signal analysis, and con-
sequently enhance the overall quality of life by enhancing disease diagnosis and
life expectancy. Further, this book endows varied communities with its innovative
advances in theory, modeling, statistical analysis, analytical approaches, analytical
results, numerical simulation, computational structuring, and case studies related to
applications of ML/DL models in the healthcare domain.

vii
Editors
Dr. Om Prakash Jena (PhD) is currently work-
ing as an Assistant Professor in the Department of
Computer Science, Ravenshaw University, Cuttack,
Odisha, India. He has ten years of teaching and
research experience in the undergraduate and post-
graduate levels. He has published several technical
papers in international journals, conferences, and
edited book chapters of reputed publications. He
is a member of IEEE, IETA, IAAC, IRED, IAENG, and WACAMLDS. His cur-
rent research interest includes database, pattern recognition, cryptography, network
security, artificial intelligence, machine learning, soft computing, natural language
processing, data science, compiler design, data analytics, and machine automation.
He has many edited books to his credit, published by Wiley, CRC Press, Bentham
Publication, and is also the author of two textbooks under Kalyani Publisher. He also
serves as a reviewer committee member and editor of many international journals.

Dr. Bharat Bhushan (PhD) is an Assistant Professor

of Department of Computer Science and Engineering
(CSE) at School of Engineering and Technology, Sharda
University, Greater Noida, India. He is an alumnus of
Birla Institute of Technology, Mesra, Ranchi, India.
He received his Undergraduate Degree (B-Tech in
Computer Science and Engineering) with Distinction
in 2012, received his Postgraduate Degree (M-Tech
in Information Security) with Distinction in 2015, and
his Doctorate Degree (Ph.D. Computer Science and
Engineering) in 2021 from Birla Institute of Technology,
Mesra, India. He has earned numerous international
certifications such as CCNA, MCTS, MCITP, RHCE,
and CCNP. In the last 3 years, he has published more
than 80 research papers in various renowned inter-
national conferences and SCI-indexed journals including Wireless Networks (Springer),
Wireless Personal Communications (Springer), Sustainable Cities and Society (Elsevier),
and Emerging Transactions on Telecommunications (Wiley). He has contributed with more
than 25 book chapters in various books and has edited 11 books from the most famed
publishers like Elsevier, IGI Global, and CRC Press. He has served as a reviewer/editorial
board member for several reputed international journals. In the past, he worked as an assis-
tant professor at HMR Institute of Technology and Management, New Delhi, and Network
Engineer in HCL Infosystems Ltd, Noida. He has passed GATE exams for successive years
and gained the highest percentile of 98.48 in GATE 2013. He is also a member of numerous
renowned bodies including IEEE, IAENG, CSTA, SCIEI, IAE, and UACEE.

ix
x Editors

Dr. Utku Kose (PhD) received the BS degree in

2008 in Computer Education at Gazi University,
Turkey as a faculty valedictorian. He received
his MS degree in 2010 from Afyon Kocatepe
University, Turkey in the field of computer and his
DS/PhD in 2017 from Selcuk University, Turkey
in the field of computer engineering. Between
2009 and 2011, he worked as a Research Assistant
in Afyon Kocatepe University. Following this, he has also worked as a Lecturer and
Vocational School Vice-Director at Afyon Kocatepe University between 2011 and
2012, as a Lecturer and Research Center Director in Usak University between 2012
and 2017, and as an Assistant Professor in Suleyman Demirel University between 2017
and 2019. Currently, he is an Associate Professor in Suleyman Demirel University,
Turkey. He is published in more than 100 publications including articles, authored
and edited books, proceedings, and reports. He is also on the editorial boards of many
scientific journals and serves as one of the editors of the Biomedical and Robotics
Healthcare book series by CRC Press. His research interest includes artificial intel-
ligence, machine ethics, artificial intelligence safety, optimization, the chaos theory,
distance education, e-learning, computer education, and computer science.
Contributors
Amira Hassan Abed Souad El Hajjaji
Department of Information Systems Faculty of Science
Center Mohammed V University
Egyptian Organization for Rabat, Morroco
Standardization & Quality
Cairo, Egypt Ahmed A. Elngar
Faculty of Computers and Artificial
C. B. Abhilash Intelligence
Indian Institute of Information Technology Beni-Suef University
Dharwad, India Beni-Suef, Egypt
K. Balakrishnan Yousef Farhaoui
Indian Institute of Information T-IDMS Faculty of Sciences and
Technology Techniques Errachidia
Tiruchirappalli, Tamil Nadu, India Moulay Ismail University
Meknes, Morocco
Shawli Bardhan
Indian Institute of Information Technology Manasi Gyanchandani
Una, India Department of CSE
MANIT
Loubna Bouhachlaf Bhopal, India
Faculty of Science
Mohammed V University Abubakar Kamagata Hamisu
Rabat, Morroco P P Savani University
Kosamba, Gujarat, India
Azedine Boulmakoul
FSTM Rajendra Hegadi
Casablanca, Morocco Indian Institute of Information Technology
Aju D Dharwad, India
Vellore Institute of Technology Hassan Homayoun
Vellore, India
Quantitative MR Imaging and
K. T. Deepak Spectroscopy Group, Research Center
Indian Institute of Information Technology for Cellular and Molecular Imaging
Dharwad, India Tehran University of Medical Sciences
Tehran, Iran
Kuldeep Dhama
Division of Pathology Om Prakash Jena
ICAR-Indian Veterinary Research Department of Computer Science
Institute Ravenshaw University
Bareilly, India Cuttack, India

R. Dhanalakshmi Sushitha Susan Joseph

Indian Institute of Information Technology Vellore Institute of Technology
Tiruchirappalli, India Vellore, India
xi
xii Contributors

Jasleen Kaur Krishna Chandra Patra

P P Savani University GIET University
Kosamba, Gujarat, India Gunupur, Odisha, India
Utkarsh Mahadeo Khaire B. Rajitha
Department of CSE Motilal Nehru National Institute of
MANIT Technology Allahabad
Bhopal, India Prayagraj, India
Nilay Khare Mithun Singh Rajput
Indian Institute of Information Technology School of Pharmacy
Dharwad, India Devi Ahilya Vishwavidyalaya
Aziz Mabrouk Indore, India
FSTM Nitin Singh Rajput
Tetouan, Morocco Vellore Institute of Technology
Jamal Mabrouki Vellore, India
Faculty of Science G. V. Eswara Rao
Mohammed V University Motilal Nehru National Institute of
Rabat, Morroco Technology Allahabad
Kavi Mahesh Prayagraj, India
Indian Institute of Information Technology Purnima Dey Sarkar
Dharwad, India Department of Medical Biochemistry
Ashima Sindhu Mohanty M.G.M. Medical College
GIET University Indore, India
Gunupur, Odisha, India
Prashant Sengar
Ranjan K. Mohapatra Indian Institute of Information
Department of Chemistry Technology
Government College of Engineering Una, India
Keonjhar, India
Essam M. Shaaban
Behzad Soleimani Neysiani Faculty of Computers & Artificial
Department of Research and Development Intelligence
Ava Aria Information Company Cairo, Egypt
Demis Holding
Isfahan, Iran Bhawna Swarnkar
Department of CSE
Madhumita Pal MANIT
Electronics Science and Engineering Bhopal, India
C. V. Raman Global University
Bhubaneswar, India Ruchi Tiwari
Department of Veterinary Microbiology
Priyadarsan Parida and Immunology
GIET University College of Veterinary Sciences
Gunupur, Odisha, India DUVASU
Smita Parija Mathura, India
Electronics Science and Engineering
C. V. Raman Global University
Bhubaneswar, India
 1 Common Data Interface
for Sustainable
Healthcare System
C. B. Abhilash, K. T. Deepak,
Rajendra Hegadi, and Kavi Mahesh

CONTENTS
1.1 Introduction....................................................................................................... 2
1.2 Related Work..................................................................................................... 3
1.2.1 Semantic Interoperability and Semantic Interoperability in EHR........ 3
1.2.2 Fast Health Interoperability Resources.................................................4
1.2.3 Existing Systems.................................................................................... 5
1.3 Terminologies in the Healthcare Ecosystem..................................................... 6
1.3.1 Standardizing Healthcare Data..............................................................6
1.3.1.1 Referencing Terminology....................................................... 6
1.3.1.2 Syntax.....................................................................................6
1.3.1.3 Semantics................................................................................7
1.3.1.4 Pragmatics...............................................................................7
1.3.2 Fast Health Interoperability Resources.................................................7
1.3.3 FHIR as API..........................................................................................7
1.3.4 Common Drug Codes for India (CDCI)................................................8
1.3.5 EHR Standards...................................................................................... 8
1.4 Methodology......................................................................................................8
1.4.1 High-Level Architecture...................................................................... 11
1.4.2 High-Level Representation of CDI Layer............................................ 11
1.4.3 Common Data Interface Functionality................................................ 12
1.4.3.1 New Patient Registration...................................................... 13
1.4.3.2 Search for an Existing Patient in Application Database....... 14
1.4.3.3 Existing Patient Information Extraction Process.................. 14
1.4.3.4 Searching for Patient Data in a Federated System................ 14
1.4.3.5 Use Case................................................................................ 15
1.4.4 CDI Subspace Creation........................................................................ 16
1.4.5 CDI ID Process and Reference Model................................................. 17
1.4.6 CDI ID Format..................................................................................... 17
1.4.6.1 Novel Unique ID Design....................................................... 17
1.5 Discussion and Conclusion.............................................................................. 19
1.6 Future Enhancement........................................................................................ 19
Acknowledgments.....................................................................................................20
References.................................................................................................................20
DOI: 10.1201/9781003226147-1 1
2 Machine Learning and Deep Learning in Medical Data Analytics

1.1 INTRODUCTION
People are well connected in more places than ever and actively participate in digi-
tal healthcare activities in the current generation. Healthcare professionals look for
deeper health insights and actionable information like making better decisions and
efficiently improving patient record information at lower costs. The current health-
care scenario is expected to be digitized. People regularly connect to their differ-
ent health gadgets and regularly monitor their health activities. Digital healthcare
enables the healthcare ecosystem to have a huge amount of connected data for
regular monitoring. The proposed high-level integration architecture is the open
architecture that provides services in compliance with data standards and has a
capability that inspires new healthcare application developers to design next-gen
connected health and wellness systems like digital gadgets that connect across
the healthcare devices, unlike other open architecture. In the current scenario,
we see a patient present at the emergency department. We don’t know anything
about them even though they might have a lot of electronic records (R. Bayer et al.
2015). Knowing them as digital citizens, we take care of them on time and at a
lesser cost. The connected healthcare system supports these features and enables
the individual’s data or health record to be accessible at any point in time for the
health ecosystem’s concerned stakeholders. So there is a necessity that automation
and integration are necessary. We have found that our automated systems are often
much siloed (T. Benson et al. 2016).
It is a long-term challenge to make the health entities interoperable. When
health data and related information are standalone systems, they need to be con-
nected to serve the purpose for which they had been originally designed or intended.
To achieve interoperability, the databases are gradually deployed in a distributed
architecture and the subsequent federation by reusing resources to build a knowl-
edge-based system. But this can be well implemented by incorporating standards
in healthcare systems. With the wide opportunity of semantic technology, we can
incorporate interoperability (B. Hu et al. 2006).
In the federated system approach, various databases are put together to exchange
and communicate the data. But this can be achieved with certain constraints in
design and usage. In a semantic-based approach, the meaning of the information
that needs to be integrated is considered when integrating the database schema with
respective row and column names. Thus, it is the syntactic approach of integration.
The amount of data in the healthcare ecosystem keeps growing. Patient data
is generated by various health stakeholders like physicians, laboratories, medical
devices, research facilities, and now even from in-home data sources such as per-
sonal fitness devices. Sharing this data across so many disparate systems is critical
to ensuring the successful care of individual patients and improving overall popula-
tion health.
In G. Alterovitz et al.’s (2015) study, the authors report a typical case system from
ontology mapping. A similar integration algorithm and data will produce differ-
ent results. The results are arbitrary, as evidenced in the formatting and annotation,
which make the result difficult to reuse. And, this makes semantic integration dif-
ficult to apply.
CDI for Sustainable Healthcare System 3

The healthcare stakeholder’s data is available in different formats, and systems

match the data generally by matching the strings. Each system has its data structure,
so it is preferred to have a standard methodology for data translation and exchange.
The standard-based translation for data interchange is recommended (T. Benson
et al., 2016). The source and target system both need to understand and interpret the
data in the same way. This chapter discusses the existing data integration techniques
available in the healthcare domain. We propose a novel methodology that intelli-
gently integrates all healthcare applications with minimal change configurations to
the existing data and structure. The proposed methodology follows the metadata
approach. A 12-digit unique ID is generated considering the respective stakehold-
ers’ predefined district and talk code, and appended by an identification code and
the unique random number. The unique ID is stored as metadata information in a
separate server for interoperable operations. The open-source tools are available to
consume or send data in fast health interoperability resources (FHIR), HL7v2 easily.
We can write code to handle data in any custom format as well. Also, the proposed
model uses different integration techniques with different functionalities as defined
in the common data interface (CDI) layer.
This paper describes developing a novel method for having unified access to
health information in the healthcare ecosystem.
This work proposes a unique CDI layer that fits the existing system architecture to
adapt the data standards to achieve interoperability. Section 1.2 illustrates the related
work considering semantics and other healthcare attributes. Section 1.3 discusses
the various healthcare data standards required to achieve interoperability. Finally, in
Section 1.4, the methodology is discussed, along with a case study considering the
patient registration process. Our proposed method uses a well-established natural
language processing technique to match the patient records with the unique ID.

1.2 RELATED WORK

In this section, the previous literature is reviewed, with regard to semantic interoper-
ability and healthcare standards. This paper aims to study the existing interoperable
approaches to enhance the health care ecosystem. With FHIR, the health data is
exchanged as resources using the XML code format. As artificial intelligence (AI)
and machine learning (ML) technologies (K. Paramesha et al. 2021) continue to
break through the restrictions of scientific drug research, ML is preferable for senti-
ment analysis (SA) for user-generated drug reviews. Many neurolinguistic program-
ming (NLP) techniques can benefit from ML’s unique learning style. Having this as
the main objective, it becomes obvious that existing literature had to be reviewed in
order to understand the process for introducing the proposed CDI layer.

1.2.1 Semantic Interoperability and Semantic Interoperability in EHR

Semantic interoperability is achieved when the data is exchanged across the inter-
faces with the required data interpretation in an unaltered way (S. Schulz et al. 2013).
Data sharing is interoperable, allowing systems to exchange information about elec-
tronic health records (EHR) and the possibility of changing the healthcare system
4 Machine Learning and Deep Learning in Medical Data Analytics

using ontology (A. Kiourtis et al. 2019). It can automatically integrate information
between multiple users and systems to improve feedback efficiency to query terms
and ensure that the feedback is true and clear regardless of the data representation
(J. D. Heflin et al. 2000). Web ontologies can be used to integrate data and seman-
tic interoperability from medical data because they use existing health standards to
access patient records. In addition, the location of data instances is consistent with
medical terminology (D. Teodoro et al. 2011).
EHR is a digital representation of a patient’s health record, including medi-
cal treatment, diagnosis, treatment plan, and medical history. This is a system-
atic way to store this information and provide it to all parties under each party’s
authorization. Therefore, EHR adopted a prototype-based approach that enables
clinical decision support system (CDSS) tools to make decisions about patient care
(R. Bayer et al. 2015).
Assuming that the EHR is patient-centric, it is not an institution. It has a long-
term care record, including the various medical care that the patient has received,
and the medical treatments, plans, and prognostic instructions followed. The EHR
prototype provides a simplified process for the flow of information between clini-
cians without interpreting information in the existing system. The terms used in the
system are not universally defined; they may be specific to a particular system devel-
oper’s specific prototype. As a result, integrating this information between different
prototypes constitutes a limitation (S. Garde et al. 2007).

1.2.2 Fast Health Interoperability Resources

The FHIR standard provides flexibility when developing ontologies by introducing
resource description framework (RDF) vocabulary to be used in framework develop-
ment. This is accomplished while preserving semantic interoperability in EHR. To
support CDS tools for precision medicine, genomic data must be linked to phenom-
ena variants in a patient’s EHR, which is done using the FHIR standard. EHR that
comply with substitutable medical applications and reusable technologies (SMART)
can be used to obtain clinical data for use in patient diagnosis, medication selection,
and care course prediction (G. Alterovitz et al. 2018).
FHIR has built-in modules like administration, clinical module, diagnostic medi-
cine, medications, and clinical reasoning, which can be used to map the ontology
framework to avoid information mismatch. The purpose of ontology has to be prop-
erly defined so that the output can be properly derived. The best approach is to base
the ontology on the FHIR standard, which has all the modules mentioned above.
For precision dosing, proper data mapping is required to achieve interoperability
(G. Alterovitz et al. 2020).
The federated health information model (FHIM) (The Open Group Healthcare
Forum 2015) is advancing the healthcare interoperability by using standards. In
this digital era, health and healthcare are privileged to be managed via technol-
ogy. However, the process of transforming healthcare information often feels
stuck in a time warp. For example, consider a case where patient X suffers from
multiple health problems like diabetes, heart disease, low back pain, obesity, and
CDI for Sustainable Healthcare System 5

depression. When patient X experiences moderate chest pain, he visits the cardi-
ologist. The concerned healthcare provider asks patient X to provide the health
conditions, insurance, and basic demographic details of patient X. This continues
as and when patient X visits all types of care providers. The problem is that while
patient X’s information is already stored at the cardiologist center, it is not acces-
sible or shareable with the diabetic care provider. Interoperability is the approach
to solve this problem to avoid delay and provide the necessary information at the
right time to the right.

1.2.3 Existing Systems
Semantic interoperability can be achieved by incorporating healthcare standards (U.
Batr et al. 2014). The author has compared various standards and their implications
to achieve semantic interoperability. Choosing the best adoptable standard is very
important. HL7 is used widely as a messaging model.
In Pijush Kanti et al.’s (2019) study the author discusses the “V’s” of healthcare
big data where volume, velocity, variability, validity, variety, veracity, viability, vul-
nerability, and visualization of data are described. To achieve the efficient use of big
data in healthcare, incorporating standards is very much required. EHR require data
across healthcare applications using standards-based methods by which seamless
data exchange can be done.
HealthSuite (Philips, USA, 2018) is a cloud-based open digital platform that offers
users continuous, personalized health care. The kit includes functions for analy-
sis, sharing, and processing. Healthcare service coordination. The analysis section
employs ML algorithms as well as various predictive analysis technologies. Shared
functionality is essentially multi-device platform interoperability. Orchestration, in
essence, achieves workflow synchronization, such as Tasks and so on (D. P. Pijush
Kanti et al. 2019).
Watson health (IBM) is a complete software package developed by IBM that
can help all aspects of health. It has AI and ML capabilities that can help diag-
nose and treat diseases effectively and reduce hospital staff and patient care staff.
Watson can understand the patient’s medical history and ask for all possible new
drugs or technologies on the market, thus saving the doctor’s time checking all the
literature.
A fully integrated system with the Internet of Healthcare Things (IoHT) frame-
work can be used remotely to assist medical experts in diagnosing and treating
skin cancer (A. Khamparia et al. 2020). According to the performance index
evaluation, the proposed framework outperforms other pre-trained architectures
regarding accuracy, recall, and accuracy of detecting and classifying skin cancer
from skin lesion images.
Even though internet of health things (IoHT) has a very complex architecture due
to the connectivity of a wide range of devices and services in the system, it can be
incorporated into the healthcare system for data collection and real-time monitoring.
This paper presents a brief overview of urban IoT systems designed to support smart
cities and advanced communication technologies (A. K. Rana et al. 2019).
6 Machine Learning and Deep Learning in Medical Data Analytics

Random Forest algorithm is a well-known decision tree-based ensemble method

that tries to increase the system accuracy and can be applied to classification and
regression applications. It has excellent data adaptability and can solve the “large p,
small n” problem. Moreover, it shows how functions interact with one another and
how they are related (P. Sudhansu Shekhar et al. 2021)
ML in healthcare can improve health information management and health
information exchange to improve work processes (P. Pattnayak et al. 2021), mod-
ernize them, make clinical data more accessible, and improve the accuracy of
health information. Most notably, it improves information processing efficiency
and transparency.
A CDSS based on an expert system will be a better solution because it will per-
form both ML functions. In addition, an expert system can assist medical staff in
diagnosing diseases when experienced doctors are unavailable in rural and remote
areas (N. Panigrahi et al. 2021). This is achieved by using interoperability.

1.3 TERMINOLOGIES IN THE HEALTHCARE ECOSYSTEM

There are several challenges when working with healthcare data. First, the data can
be transmitted in any number of formats. For example, HL7v2 or some completely
customized format. Second, sharing the data in different formats between systems
and integrating them is difficult and time-consuming.
The second challenge is that, even if the data is in the same format, it is repre-
sented differently by the various systems that use this data. There are several coding
systems such as LONIC, SNOMED, ICD-10. So sharing results that are coded dif-
ferently within systems is challenging since translation is required to integrate. It is
necessary to have one standard, or a mapping engine that works between standards
should be developed.

1.3.1 Standardizing Healthcare Data

S. Schulz et al. (2019) seek classification in a report titled “Medical Data Standards.”
They clarify relevant, language-driven concepts to define the types of data standards
in healthcare. Also, the authors describe four concepts that characterize various
aspects of clinical data, as outlined below (ISO/TR 20514:2005).

1.3.1.1 Referencing Terminology

Referencing terminology is the set of unique, human-understandable, unambiguous,
standardized labels for terms is referred to as a reference term that is, the term used
as a reference.

1.3.1.2 Syntax
Syntax is the standard that specifies the required order of composition when ana-
tomical terms and various restrictions, such as “acute,” “distal,” “left and right,” are
used in clinical narratives.