See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/331302149

Data Science for Healthcare: Methodologies and Applications

Book · January 2019

DOI: 10.1007/978-3-030-05249-2

CITATIONS READS

0 2,700

3 authors, including:

Sergio Consoli
European Commission
83 PUBLICATIONS   480 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Refugium Peccatorum View project

Linked Open Data for Public Administrations View project

All content following this page was uploaded by Sergio Consoli on 25 February 2019.

The user has requested enhancement of the downloaded file.

Data Science for Healthcare
Sergio Consoli • Diego Reforgiato Recupero •
Milan Petković
Editors

Data Science for Healthcare

Methodologies and Applications

123
Editors
Sergio Consoli Diego Reforgiato Recupero
Philips Research Dept of Mathematics and Computer Science
Eindhoven, The Netherlands University of Cagliari
Cagliari, Italy

Milan Petković
Data Science Department
Philips Research
Eindhoven, The Netherlands

ISBN 978-3-030-05248-5 ISBN 978-3-030-05249-2 (eBook)

https://doi.org/10.1007/978-3-030-05249-2

Library of Congress Control Number: 2018966867

© Springer Nature Switzerland AG 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Foreword

It is becoming obvious that only by fundamentally rethinking our healthcare systems

we can successfully address the serious challenges we are facing globally.
One of the most significant challenges is the aging of populations, which comes
with a high percentage of chronically ill people, often with multiple conditions. In
addition, there is a rising incidence of preventable lifestyle-related diseases caused
by risk factors such as obesity, smoking, and alcohol consumption. Today, chronic
diseases in EU already result in the loss of 3.4 million potential productive life years,
which amounts to an annual loss of e115 billion for the EU economy. At the same
time, we are being faced with a shortage of qualified healthcare professionals, and
with quality and efficiency issues in the way healthcare is delivered. Finally, public
spending on healthcare is steadily rising. The EU spends around 10% of its GDP
on healthcare. In 2015, US healthcare spending increased 5.8% to $3.2 trillion. The
costs are expected to continue rising—to unaffordable levels.
We need to transition to new care delivery models, addressing the quadruple
aim of (1) improving the health of populations, (2) reducing the per capita cost of
healthcare, (3) improving the patient experience including quality and satisfaction,
and (4) improving the work life of healthcare providers by providing necessary
support.
The good news is that digital technologies are by now so powerful, affordable,
and pervasive, that they help to make these goals achievable. The Internet of Medical
Things and artificial intelligence (AI) in particular are key enablers of the digital
transformation in healthcare. Connected medical devices will soon be everywhere,
from hospital to home, providing a rich variety of data. AI will be instrumental in
turning these data into actionable insights across the continuum of care.
But technology by itself will not be the answer. In the end, healthcare is all about
people. Meaningful innovation occurs when technology enables professionals to
deliver better care and when it empowers consumers and patients to better manage
their own health. This means that applying AI and data science to healthcare requires
a deep understanding of the personal, clinical, or operational context in which they
are used. That is why, at Philips, we believe in the power of adaptive intelligence.

v
vi Foreword

Adaptive intelligence combines AI with human domain knowledge to create

solutions that adapt to people’s needs and environments—supporting them in their
daily work and lives. Adaptive intelligence augments people, rather than replacing
them. It acts like a personal assistant that can learn and adapt to the skills and
preferences of the person that uses it, and to the situation he or she is in. The
technology does not call attention to itself, but runs in the background—deeply
integrated into the interfaces and workflows of hospitals, and almost invisibly
embedded into solutions for the consumer environment.
This is not merely a future vision—it is becoming a reality today. This book
includes examples that show how data science and AI-enabled solutions are
already supporting clinical care and prevention of disease or health incidents. It
is very encouraging that advances in AI methods such as machine learning, natural
language processing, and computer vision can all improve people’s lives, when they
are employed wisely.
As we continue to make strides in the digital transformation of healthcare
systems, it is important to be aware of the possibilities of AI and data science—
and how they can be used in an effective and responsible way to help achieve
the quadruple aim. This book will help the reader to learn how to (1) extract new
knowledge from health data to improve healthcare delivery, (2) enable healthcare
systems to deliver better outcomes at lower costs, and (3) support the transition
from an acute, episodic care model to proactive chronic disease management.
Enjoy the read, and join this exciting journey!

Chief Technology Officer, Philips Henk van Houten

Eindhoven, The Netherlands
Preface

Healthcare systems around the world are facing vast challenges in responding to
trends of aging population, the rise of chronic diseases, resources constraints, and
the growing focus of citizens on healthy living and prevention. Consequently, there
is an increasing focus on answering important questions such as: (1) How do we
improve the rate of fast, accurate first-time-right diagnoses? (2) How can we reduce
the huge variance in costs and outcomes in health systems? (3) How do we get
people to take more accountability for their own health? (4) How can we provide
better health care at lower cost?
On the other hand, digitization and rapid advances in ICT technology are
enabling the capture of more data than ever before, including medical health records,
people’s vital signs and their lifestyle, data about health systems, and data about
population health in general. This tsunami of data per se does not immediately result
in better healthcare insights, but, on the contrary, if not used properly, it can be a
burden to people and result in clinicians spending more time with computers than
face to face with patients, or citizens being lost in data they are getting from health
trackers and many different sensors, or, again, patients reluctant to accept assistive
technologies. This is exactly the point where unlocking the power of data science
and artificial intelligence can help by making sense of the large amounts of data,
turning them into actionable insights providing mutual benefits to both patient and
medical professionals, also helping in answering the abovementioned questions.

Aim

The goal of this book is to boost the adoption of data science and artificial
intelligence solutions for healthcare by raising awareness of existing proof points
of these applications and underlying world-class innovations on data science and
artificial intelligence in healthcare. The book builds on several interconnected
disciplines, including advanced machine learning, big data analytics, data mining,
statistics, probabilistic modeling, pattern recognition, computer vision, and seman-

vii
viii Preface

tic reasoning, with direct application to modern HealthTech. Consequently, it shows

how the advances in the aforementioned scientific disciplines, as well as digital
data platforms, can create value within the healthcare domain and help in reaching
the quadruple aim of improving healthcare outcomes, lowering the cost of care,
enhancing the patient experience, and improving the work life of care providers.
In particular, the focus of this book is threefold. Firstly, the book aims at
demystifying data science and artificial intelligence methods that can be used to
extract new knowledge from health data and to improve healthcare delivery. The
application of digital technologies for healthcare is seeing a gradual transition to
integrated care delivery networks with the consumer at the center. The incoming
trends include increased self-management and individualized treatment paths. Thus,
secondly, the focus is on applications that enable health systems to deliver better
outcomes at lower cost, by boosting the digitization of the healthcare system. This
is the starting point for the application of data science and artificial intelligence
technologies supporting the move from reactive acute care to pro-active chronic
disease management, which is the third focus point of this book. By unlocking the
power of big data, connected health systems will be able to deliver personalized and
industrialized care models that will lead to a new era of outcome-based healthcare.

Organization

The book starts with three solid tutorial chapters on data science in healthcare,
to help readers understand the opportunities and challenges; become familiar
with the latest methodological findings in machine learning, in particular deep
learning, for healthcare; and help them understand how to use and evaluate the
performance of novel data science and artificial intelligence tools and frameworks.
These chapters are followed by 11 other chapters showing successful stories on the
application of the specific data science technologies in healthcare. The discussed
data science technologies and their applications in healthcare focus on, among
others, supervised learning, unsupervised learning, deep learning, natural language
processing, information retrieval, knowledge management and reasoning, data-
to-text, cognitive computation, process mining, smart networking, computational
optimization, visual analytics, and robotics.

Audience

This book is primarily intended for data scientists involved in the healthcare domain.
There is a clear need for healthcare data analysts to make sense of clinical and
personally generated health data more systematically. By reading this book, on
one hand computer scientists involved in the medical sector will be able to learn
the modern effective data science technologies to create innovation for HealthTech
Preface ix

businesses; on the other, experts involved in the healthcare sector will become more
familiar with the advances in ICT and will be able to analyze and process (big) data
in order to apply these technologies holistically for patient care. Prior knowledge in
data science with real-world applications to the healthcare sector is recommended
to interested readers in order to have a clear understanding of this book.

Final Words

We are quite convinced that artificial intelligence and data science will further
advance, creating a great potential to industrialize the healthcare sector and to
improve the quality of healthcare while managing the costs. In the long run, these
technologies might be so impactful that they could result in a giant leap of humanity,
changing also the healthcare beyond our current expectations and bringing it closer
to maintenance of robotic technology. Let’s see which future we will create. Enjoy
the reading!

Eindhoven, The Netherlands Sergio Consoli

Cagliari, Italy Diego Reforgiato Recupero
Eindhoven, The Netherlands Milan Petković
Contents

Part I Challenges and Basic Technologies

Data Science in Healthcare: Benefits, Challenges and Opportunities .. . . . . 3
Ziawasch Abedjan, Nozha Boujemaa, Stuart Campbell, Patricia Casla,
Supriyo Chatterjea, Sergio Consoli, Cristobal Costa-Soria, Paul Czech,
Marija Despenic, Chiara Garattini, Dirk Hamelinck, Adrienne Heinrich,
Wessel Kraaij, Jacek Kustra, Aizea Lojo, Marga Martin Sanchez,
Miguel A. Mayer, Matteo Melideo, Ernestina Menasalvas,
Frank Moller Aarestrup, Elvira Narro Artigot, Milan Petković,
Diego Reforgiato Recupero, Alejandro Rodriguez Gonzalez,
Gisele Roesems Kerremans, Roland Roller, Mario Romao, Stefan Ruping,
Felix Sasaki, Wouter Spek, Nenad Stojanovic, Jack Thoms,
Andrejs Vasiljevs, Wilfried Verachtert, and Roel Wuyts
Introduction to Classification Algorithms and Their Performance
Analysis Using Medical Examples. . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 39
Jan Korst, Verus Pronk, Mauro Barbieri, and Sergio Consoli
The Role of Deep Learning in Improving Healthcare.. . .. . . . . . . . . . . . . . . . . . . . 75
Stefan Thaler and Vlado Menkovski

Part II Specific Technologies and Applications

Making Effective Use of Healthcare Data Using Data-to-Text
Technology . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 119
Steffen Pauws, Albert Gatt, Emiel Krahmer, and Ehud Reiter
Clinical Natural Language Processing with Deep Learning.. . . . . . . . . . . . . . . . 147
Sadid A. Hasan and Oladimeji Farri

xi
 xii Contents

Ontology-Based Knowledge Management for Comprehensive

Geriatric Assessment and Reminiscence Therapy on Social Robots . . . . . . . 173
Luigi Asprino, Aldo Gangemi, Andrea Giovanni Nuzzolese,
Valentina Presutti, Diego Reforgiato Recupero, and Alessandro Russo
Assistive Robots for the Elderly: Innovative Tools to Gather Health
Relevant Data .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 195
Alessandra Vitanza, Grazia D’Onofrio, Francesco Ricciardi,
Daniele Sancarlo, Antonio Greco, and Francesco Giuliani
Overview of Data Linkage Methods for Integrating Separate
Health Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 217
Ana Kostadinovska, Muhammad Asim, Daniel Pletea, and Steffen Pauws
A Flexible Knowledge-Based Architecture for Supporting
the Adoption of Healthy Lifestyles with Persuasive Dialogs . . . . . . . . . . . . . . . . 239
Mauro Dragoni, Tania Bailoni, Rosa Maimone, Michele Marchesoni,
and Claudio Eccher
Visual Analytics for Classifier Construction and Evaluation
for Medical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 267
Jacek Kustra and Alexandru Telea
Data Visualization in Clinical Practice . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 289
Monique Hendriks, Charalampos Xanthopoulakis, Pieter Vos,
Sergio Consoli, and Jacek Kustra
Using Process Analytics to Improve Healthcare Processes . . . . . . . . . . . . . . . . . . 305
Bart Hompes, Prabhakar Dixit, and Joos Buijs
A Multi-Scale Computational Approach to Understanding Cancer
Metabolism. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 327
Angelo Lucia and Peter A. DiMaggio
Leveraging Financial Analytics for Healthcare Organizations
in Value-Based Care Environments . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 347
Dieter Van de Craen, Daniele De Massari, Tobias Wirth, Jason Gwizdala,
and Steffen Pauws

View publication stats