![Methods in Biomedical Informatics.
© 2014 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/B978-0-12-401678-1.00001-4
1
Chapter
1.1
BIOMEDICAL INFORMATICS AND ITS
APPLICATIONS
Biomedicine is a complex ecosystem of foundational and applied disciplines.
Generally, researchers and practitioners in biomedicine are specialized in a
particular area of emphasis that can be described as either: (1) “bench”—
aiming to understand the underpinning principles of dysfunction leading to
compromised health status; (2) “bedside”—aiming to develop preventative
approaches for the maintenance of normal health status or addressing abnor-
mal health status through clinical intervention and quantifying the effect on
individuals; or (3) “community”—aiming to quantify the effect of preventa-
tive measures or treatment regimens across defined population groups and
developing systematic approaches to promote effective practices. There has
been increased discussion in recent years about these areas of emphasis and
the need for strategies to promote synergies between them [1–3].
This book is focused on describing methodologies that underpin the disci-
pline of Biomedical Informatics, which is defined as a field “that studies and
pursues the effective uses of biomedical data, information, and knowledge
for scientific inquiry, problem solving and decision making, motivated by
efforts to improve human health [4].” Biomedical informatics has evolved
into a discipline that encompasses the bench, bedside, and community areas
of biomedicine and represents a synergy between foundational and applied
aspects of biomedical science [4]. The foundational aspects of biomedi-
cal informatics are derived from a wide array of fundamental disciplines,
1
CHAPTER OUTLINE
1.1 Biomedical Informatics and its Applications 1
1.2 The Scientific Method 4
1.3 Data, Information, Knowledge, and Wisdom 6
1.4 Overview of Chapters 9
1.5 Expectations and Challenge to the Reader 11
References 12
Introduction
Indra Neil Sarkar
Center for Clinical and Translational Science, University of Vermont, Burlington, VT, USA](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-11-320.jpg)
![2 CHAPTER 1 Introduction
including those from formal science (e.g., logic, mathematics, and statistics)
and basic science (e.g., physics, chemistry, biology, psychology, and eco-
nomics). Additional applied disciplines also contribute key foundational
elements to biomedical informatics (e.g., medical science, computer sci-
ence, engineering, and epidemiology).
From the foundational sciences that contribute to biomedical informatics,
methodologies are developed and applied to specific biomedical areas.
For example, biomedical informatics methods may be applied to spe-
cific domain areas, like: biology (termed “bioinformatics”), medicine and
nursing (termed “clinical informatics”), and public health (termed “pub-
lic health informatics”). In contemporary biomedical informatics par-
lance, “health informatics” is often used as an umbrella term for clinical
informatics and public health informatics. There are many other areas of
informatics that traverse domain boundaries (e.g., “imaging informatics”
involves the application of biomedical informatics approaches to areas that
span biological and clinical domains) or are aligned with particular spe-
cialties (e.g., “pathology informatics”). A comprehensive survey of bio-
medical informatics and its application areas can be found in Biomedical
Informatics: Computer Applications in Health Care (edited by Shortliffe
and Cimino) [5].
Biomedical informatics methods may also be applied to areas that focus on
the translation of innovations from one part of the biomedicine spectrum
to another [6]. In recent years, this area of emphasis has collectively been
referred to as “translational sciences.” Working alongside translational med-
icine, the translational sciences aim to enable greater synergy between basic
and applied sciences [7,8]. Within the context of biomedical informatics,
“translational informatics” is the use of informatics approaches to support
translational science and medicine [9]. Translational informatics is divided
into two major sub-specialties: (1) “translational bioinformatics”—which
is focused on the development of informatics approaches that facilitate the
translation of basic science findings into ones that may be utilized in clinical
contexts [10,11]; and (2) “clinical research informatics”—which is focused
on the development of informatics approaches for evaluating and gener-
alizing clinical innovations [12]. Translational bioinformatics and clinical
research informatics work in tandem to increase the likelihood of basic sci-
ence innovations being put into clinical practice and improving the health of
communities—thus directly addressing the translational bottlenecks, such
as those that are often seen at the transition of knowledge from bench to
bedside or bedside to community. Figure 1.1 provides a graphical overview
biomedical informatics and its sub-disciplines that span the full breath of
biomedicine.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-12-320.jpg)
![3
1.1 BiomedicalInformaticsanditsApplications
The expansiveness of biomedical informatics in many ways requires a poly-
mathic approach to developing approaches for the betterment of health.
Formal training in biomedical informatics thus requires a unique combina-
tion of formal, basic, and applied science [4]. It therefore is not uncommon
for one to first attain formal, graduate-level scholarly or professional train-
ing in at least one area of science before embarking on additional training in
biomedical informatics. In addition to these cross-trained individuals, there
are also a growing number of formally trained biomedical informaticians
whose entire graduate-level education is done in biomedical informatics.
In most cases, biomedical informaticians choose at least one application
area of specialization (e.g., bioinformatics, clinical informatics, or public
health informatics). Regardless of the path to becoming a biomedical infor-
matician, the approaches used to address biomedical problems are built on a
common set of methodologies [4]. It must be noted that the highly special-
ized training required for success in biomedical informatics has resulted in a
significant shortage of biomedical informaticians across the entire spectrum
of biomedicine. To address this challenge, there are an increasingly grow-
ing number of formal training opportunities that strive to help provide the
biomedical enterprise with biomedical informaticians [13].
In contrast to multi- or inter-disciplinary disciplines, where foundational ele-
ments are respectively combined in additive or interactive ways, biomedical
informatics is a trans-disciplinary discipline, where foundational elements
are holistically combined in ways that result in the emergence of entirely
new concepts [14–16]. To this end, biomedical informatics is a unique dis-
cipline in that it brings together an array of diverse expertise and experi-
ence, but remains with the singular purpose to develop methods to improve
the process of health maintenance and treatment of deviations from normal
Bench Bedside
Bioinformatics Clinical Informatics
Community
Public Health Informatics
Biomedical Informatics
Health Informatics
Translational Bioinformatics Clinical Research Informatics
Translational
Sciences
Domain
Sciences
Translational Informatics
n FIGURE 1.1 Overview of Biomedical Informatics. Major application areas of biomedical informat-
ics are shown, aligned according to their general emphasis to develop solutions for bench, bedside, or
community stakeholders. The major application areas are also segregated according to their domain or
translational science centricity. Note that not all application areas are shown.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-13-320.jpg)
![4 CHAPTER 1 Introduction
health. The development and use of these methodologies can be organized
according to the Scientific Method (described in Section 1.2), with the goal
of transforming biomedical data into information that leads to actionable
knowledge and therefore leading to wisdom (collectively referred to as the
DIKW framework, described in Section 1.3). This book provides an intro-
duction to a number of key methodologies used in biomedical informatics
that hone the Scientific Method in the context of the DIKW framework (an
overview of the chapters is provided in Section 1.4, along with expectations
in Section 1.5).
1.2
THE SCIENTIFIC METHOD
Like all scientific disciplines, biomedical informatics is strongly grounded
in the Scientific Method. The Scientific Method can be traced back to
Artistotle, who introduced the principles of logic coming in two forms [17]:
(1) inductive—which makes postulations based on observation of universal
concepts; and (2) deductive—which makes postulations based on relation-
ships between already accepted universal concepts (called “syllogisms”). It
was not, however, until the Arab polymath al-Hasan ibn al-Haytham (often
referred to in the Western world as “Alhazen” or “Alhacen”) described the
principles of optics in a systematic manner that the contemporary Scientific
Method became a formally described process [18]. Ibn al-Haytham’s Book
of Optics was one of the main sources used by the Englishman Roger Bacon
(not to be confused with Francis Bacon, who described an alternative induc-
tive methodology referred to as the “Baconian Method,” which in many
ways rejected the hypothesis-driven Scientific Method [19]) to formally
describe the Scientific Method to the Western world [20].
The Scientific Method consists of five major activities: (1) Question for-
mulation; (2) Hypothesis Generation; (3) Prediction; (4) Testing; and (5)
Analysis. The first activity of question formulation aims to identify a query
of interest relative to a specific observation (e.g., “will this treatment regi-
men cure the patient of this illness?”). Question formulation can involve the
consultation of existing knowledge sources or other experts for determining
the validity of the question. Question formulation, in many ways, is the most
difficult step of the Scientific Method; however, it is also the most crucial
step because all the consequent steps are dependent on a well-formed ques-
tion. Once a question is developed, the next stage of the Scientific Method
strives to add focus by postulating a specific hypothesis (e.g., “this treatment
will treat the patient of their illness.”).
The generation of a hypothesis is often done in two parts, which together
make the hypothesis testable: (1) defining the null hypothesis (H0)—a con-
templation of the hypothesis in a statistical framework that presents the](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-14-320.jpg)
![5
1.2 TheScientificMethod
default conclusion (e.g., “this treatment regimen does not cure the patient
of their illness compared to a placebo.”); and (2) defining the alternative
hypothesis (H1)—a contemplation of the hypothesis in a statistical frame-
work that presents the desired outcome (e.g., “this treatment, when com-
pared to a placebo, cures the patient of their illness”). An important feature
of a well-formed hypothesis is that it is falsifiable. Falsifiability is defined as
the ability to identify potential solutions that could disprove a stated hypoth-
esis (e.g., “the patient is cured by a placebo.”). Thus, the testability of a
hypothesis is inherently a feature of its falsifiability [21].
After it is deemed that a hypothesis is testable, the next stage in the Scientific
Method is to propose some predictions that help determine the plausibility
of the hypothesis relative to alternatives, including coincidence (e.g., “this
treatment cures the patient of their illness compared to doing nothing”). The
predictions are then tested through the gathering of evidence, which support
or refute the hypothesis of interest (e.g., “for a sample of chosen patients
who have the same illness, measure the effect of the treatment versus a pla-
cebo versus nothing”). As noted earlier, an important feature of developed
tests is that they aim to address the falsifiability of the hypothesis.
The results of the tests are then analyzed to determine if the hypothesis was
indeed proved true (rejection of the null hypothesis and thus acceptance of
the alternative hypothesis). This is commonly done using a statistical com-
parison test such as one that can be derived from a confusion matrix [which
delineates verifiable (“true positive” and “true negative”) and unexpected
(“false positive” and “false negative”) results based on previous knowl-
edge]. Common comparison tests include the Pearson’s chi-square [22] and
Fisher’s exact test [23]. The final outcome of the analysis is a statement
relative to the original question (e.g., “yes, this treatment regimen will cure
patients of this illness.”).
The Scientific Method does not necessarily conclude at the completion
of the analysis step; the analysis of one hypothesis may lead to additional
questions that can consequently be examined through an additional itera-
tion of the Scientific Method. Indeed, the Scientific Method as classically
implemented can be perceived as an infinite process. Within the context of
biomedicine, a modern interpretation of the Scientific Method is often used,
termed the “hypothetico-deductive approach.” The hypothetico-deductive
approach is a cornerstone in clinical education and practice [24–26]: (1)
gather data about a patient; (2) develop questions that lead to hypotheses for
the patient’s state; (3) propose predictions based on the suggested hypotheses
that explain the patient’s state; and (4) test the hypotheses through attempts
at falsifying them to explain the patient’s state. It is important to once again](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-15-320.jpg)
![6 CHAPTER 1 Introduction
underscore that clinical inquiry into patient status is not a verification of a
particular condition; evidence to support a proposed set of hypotheses is
done through a series of falsification tests (in clinical parlance, these are
often termed “rule-in” or “rule-out” tests). As in many clinical scenarios, it
must be acknowledged that the absolute true diagnosis for a given patient
may not be known or even knowable (e.g., to determine whether a patient
actually has Alzheimer disease, the most definitive method of detection is a
neuropathological analysis done post-mortem [27]); however, the develop-
ment of hypotheses and tests to prove or refute them enables a clinician to
get closer to the true diagnosis.
1.3
DATA, INFORMATION, KNOWLEDGE, AND
WISDOM
Data are composed of individual datum points that may originate from a
plethora of sources (it is for this reason that in scientific writing the term
“data” should be treated as a plural noun; the singular form is “datum”).
Simply put, data are the raw substrate that can be transformed through for-
mal science methods to acquire meaning. Data may come in many forms,
from any part of the biomedical spectrum. Data are the essential build-
ing blocks wherefrom the Scientific Method begins and may lead to the
gathering of additional data in the testing of hypotheses. In biomedicine,
data are generated from a wide range of sources—as artifacts of digital
systems (e.g., as might be generated from automated laboratory systems)
or as recorded events from human-human interactions (e.g., as might be
generated from a clinician-patient interview). Data are an essential com-
ponent of modern science, and can acquire different meanings depending
on their interpretation. These interpretations are dependent on the use of
appropriate formal science techniques—most often a combination of logic
and statistics.
It is important for one to be aware of the assumptions made in both the
logic (e.g., the definition of true versus false) and statistics (e.g., the data
distributed in a Gaussian manner). An additional layer of complexity is that
data can, and often do, suffer from unavoidable inconsistencies, which may
be an artifact of either the generation, collection, or interpretation of data.
Even with all these issues, data form the basis for all derived interpretations
and thus form the foundation for the biomedical sciences. It is the study
and improvement of formal science techniques that also form the basis of
biomedical informatics methods.
As data are transformed into information, which is the result of the appli-
cation of formal science techniques, hypotheses may be generated about](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-16-320.jpg)
![7
1.3 Data,Information,Knowledge,andWisdom
the meaning of the observed data that form the basis for basic science.
The basic sciences, which have a strong basis in the Scientific Method
(built around the generation, testing, and validation of hypotheses),
impart interpretations of data based on a specific domain focus. For
example, physics is focused on the interpretation of data to understand
the physical world in terms of matter and its motion; chemistry is focused
on analyzing data to understand the composition of matter; biology is
focused on understanding extinct and extant organisms; and economics
is focused on the understanding of data associated with goods or services
that form the basis for inter-personal relationships. The basic sciences
form the foundation for biomedicine, and provide the insights about the
underlying cause of dysfunction and its detection as well as provide the
tools for tracking treatment outcomes from both clinical and economical
perspectives.
As information about health is collected and analyzed, it can be coalesced
into reusable constructs (e.g., a particular treatment regimen that has been
shown to improve health). These constructs form the basis of knowledge,
which require a systematic understanding and use of data that have been
transformed into interpretable information. Knowledge can be used to guide
decisions or subsequent analyses—this type of knowledge is referred to as
“actionable knowledge [28].” Actionable knowledge is of the most utility in
biomedicine when data can be used in a way to guide clinical decisions in a
manner to positively affect patient health outcomes. The study and evalua-
tion of knowledge leads to wisdom, which promotes the development of best
practices and guidance for how to interpret future encounters with biomedi-
cal data.
The applied sciences provide the scaffolding to systematically transform
information from biomedical data into knowledge and retain it as wisdom
that can be used to guide disease diagnoses, treatment regimens, or out-
comes analyses. For example, medical science approaches can be used to
consistently interpret a laboratory result in combination with a collection of
signs and symptoms to determine the health status of a patient; engineering
approaches can be used to develop a process to provide consistent levels of
care to patients as they interact with the health care system; and epidemi-
ology approaches can be used to analyze the effect of vaccinations across
susceptible populations.
Data, information, knowledge, and wisdom and their relationships to each
other are collectively referred to as the Data-Information-Knowledge-
Wisdom (DIKW) framework, and the most common interpretations are cred-
ited to Ackoff [29, 30]. The DIKW framework (where data are transformed](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-17-320.jpg)
![8 CHAPTER 1 Introduction
into wisdom) can certainly be applied to each individual scientific disci-
pline; however, as noted earlier, what sets biomedical informatics apart from
other disciplines is that its scope is trans-disciplinary. The DIKW frame-
work can be applied in specific formal, basic, or applied science contexts;
however, in the context of biomedical informatics, the DIKW framework is
used to bridge formal, basic, and applied sciences toward a single purpose—
to improve the diagnosis, care, and treatment of illness. Furthermore, the
DIKW framework in biomedical informatics formalizes the implementa-
tion of the Scientific Method towards the discovery and implementation of
biomedical innovations. Like the Scientific Method, the DIKW framework
does not necessarily conclude at the establishment of wisdom; the wisdom
gained from one set of data can be used as data for a subsequent study. The
relationships between the sciences, the Scientific Method, and the DIKW
framework are graphically depicted in Figure 1.2.
The DIKW framework can be used to formally organize, study, and innovate
at different levels of inquiry. Put into the context of biomedicine, the DIKW
framework is an essential process that transverses the aforementioned major
areas of bench, bedside, and community. The results of transforming data
into wisdom in one area may very well lead to data for another area. Of
course, the boundaries between the suggested areas of biomedicine are not
always apparent. Recently, there has been a concerted effort in the biomedi-
cal research enterprise to acknowledge that much of biomedicine innovation
does suffer from a “siloed” approach that force the bench, bedside, and com-
munity researchers into disjointed endeavors [1–3].
From the earliest days of biomedical informatics, it was proposed that con-
cepts and approaches, such as those that could be implemented in computers,
could be used to better integrate the bench, bedside, and community areas
of research and practice [31–33]. Biomedical informaticians thus necessarily
Data Information Knowledge Wisdom
Scientific Method
Question
Formulation
Hypothesis
Generation
Prediction Testing Analysis
e.g., Mathematics, Logic,
Statistics
Formal Sciences Basic Sciences
e.g., Physics, Chemistry,
Biology, Economics
Applied Sciences
e.g. Medical Science, Engineering,
Epidemiology
n FIGURE 1.2 DIKW Framework. The Data-Information-Knowledge-Wisdom (DIKW) framework is
depicted as a process that unifies the formal, basic, and applied sciences. The graphic also shows how
the major steps of the Scientific Method can be aligned with the transformation of data into wisdom.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-18-320.jpg)
![12 CHAPTER 1 Introduction
For this book to be used as an effective educational instrument or as an intro-
duction to the broad range of methodologies that are used in biomedical infor-
matics, it must be used as a starting point and not as a comprehensive reference
for a given methodology. Many of the chapter authors have included references
that may be consulted for additional details. Whilst a book of this nature will
never be comprehensive (or even complete in topics covered), it is expected
that it will provide a foundation for the methodological principles of biomedi-
cal informatics that can be applied across the spectrum of biomedicine.
REFERENCES
[1] Keramaris NC, Kanakaris NK, Tzioupis C, Kontakis G, Giannoudis PV. Translational
research: from benchside to bedside. Injury 2008;39(6):643–50.
[2] Woolf SH. The meaning of translational research and why it matters. J Am Med
Assoc 2008;299(2):211–3.
[3] Westfall JM, Mold J, Fagnan L. Practice-based research–“Blue Highways” on the
NIH roadmap. J Am Med Assoc 2007;297(4):403–6.
[4] Kulikowski CA, Shortliffe EH, Currie LM, Elkin PL, Hunter LE, Johnson TR, et al.
AMIA board white paper: definition of biomedical informatics and specification of
core competencies for graduate education in the discipline. J Am Med Inform Assoc
2012;19(6):931–8.
[5] Shortliffe EH, Cimino JJ, editors. Biomedical informatics: computer applications in
health care and biomedicine. 4th ed. New York: Springer; 2013.
[6] Sarkar IN. Biomedical informatics and translational medicine. J Transl Med
2010;8:22. PubMed PMID: 20187952.
[7] Lean ME, Mann JI, Hoek JA, Elliot RM, Schofield G. Translational research. BMJ
2008;337:a863.
[8] Wehling M. Principles of translational science in medicine: from bench to bedside.
Cambridge, New York: Cambridge University Press; 2010. xxii, 382 p., 24 p. of
plates p.
[9] Payne PR, Embi PJ, Sen CK. Translational informatics: enabling high-throughput
research paradigms. Physiol Genomics 2009;39(3):131–40.
[10] Altman RB. Translational bioinformatics: linking the molecular world to the clinical
world. Clin Pharmacol Ther 2012;91(6):994–1000.
[11] Sarkar IN, Butte AJ, LussierYA, Tarczy-Hornoch P, Ohno-Machado L. Translational
bioinformatics: linking knowledge across biological and clinical realms. J Am Med
Inform Assoc 2011;18(4):354–7.
[12] Embi PJ, Payne PR. Clinical research informatics: challenges, opportunities and defi-
nition for an emerging domain. J Am Med Inform Assoc 2009;16(3):316–27.
[13] Shortliffe EH. The future of biomedical informatics: a perspective from academia.
Stud Health Technol Inform 2012;180:19–24.
[14] Choi BC, Pak AW. Multidisciplinarity, interdisciplinarity and transdisciplinarity
in health research, services, education and policy: 1. Definitions, objectives, and
evidence of effectiveness. Clin Invest Med. Medecine clinique et experimentale
2006;29(6):351–64.
[15] Choi BC, Pak AW. Multidisciplinarity, interdisciplinarity, and transdisciplinarity
in health research, services, education and policy: 2. Promotors, barriers, and](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-22-320.jpg)
![13
References
strategies of enhancement. Clin Invest Med. Medecine clinique et experimentale
2007;30(6):E224–32.
[16] Choi BC, Pak AW. Multidisciplinarity, interdisciplinarity, and transdisciplinarity in
health research, services, education and policy: 3. Discipline, inter-discipline dis-
tance, and selection of discipline. Clin Invest Med. Medecine clinique et experimen-
tale 2008;31(1):E41–8.
[17] Barnes J. The Cambridge companion to Aristotle. Cambridge, NewYork: Cambridge
University Press; 1995. xxv, p. 404.
[18] Omar SB. Ibn al-Haytham’s optics: a study of the origins of experimental science.
Minneapolis: Bibliotheca Islamica; 1977. p. 168.
[19] Bacon F, Jardine L, Silverthorne M. The new organon. Cambridge U.K., New York:
Cambridge University Press; 2000. xxxv, p. 252.
[20] Hackett J. Roger Bacon and the sciences: commemorative essays. Leiden, NewYork:
Brill; 1997. x, p. 439.
[21] Popper KR. The logic of scientific discovery. London, New York: Routledge; 1992.
p. 479.
[22] Pearson K. On the criterion that a given system of deviations from the probable in the
case of a correlated system of variables is such that it can be reasonably supposed to
have arisen from random sampling. Philos Mag Ser 5 1900;50(302):157–75.
[23] Fisher RA. On the interpretation of χ2
from contingency tables, and the calculation
of P. J Roy Stat Soc 1922;85(1):87–94.
[24] Barrows HS, Tamblyn RM. Problem-based learning: an approach to medical educa-
tion. New York: Springer Pub. Co.; 1980. xvii, p. 206.
[25] Mandin H, Jones A, Woloschuk W, Harasym P. Helping students learn to think
like experts when solving clinical problems. Acad Med: J Assoc Am Med Coll
1997;72(3):173–9.
[26] Connelly DP, Johnson PE. The medical problem solving process. Hum Pathol
1980;11(5):412–9.
[27] Beach TG, Monsell SE, Phillips LE, Kukull W. Accuracy of the clinical diagno-
sis of Alzheimer disease at National Institute on Aging Alzheimer Disease Centers,
2005–2010. J Neuropathol Exp Neurol 2012;71(4):266–73.
[28] Cao L. Domain driven data mining. New York, London: Springer; 2010. xvi, p. 248
[29] Ackoff RL. From data to wisdom. J Appl Syst Anal 1989;16:3–9.
[30] Rowley J. The wisdom hierarchy: representations of the DIKW hierarchy. J Inf Sci
2007;33(2):163–80.
[31] Ledley RS, Lusted LB. Reasoning foundations of medical diagnosis; symbolic
logic, probability, and value theory aid our understanding of how physicians reason.
Science 1959;130(3366):9–21.
[32] Lusted LB, Ledley RS. Mathematical models in medical diagnosis. J Med Educ
1960;35:214–22.
[33] Blois MS. Information holds medicine together. MD Comput: Comput Med Practice
1987;4(5):42–6.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-23-320.jpg)
![http://dx.doi.org/10.1016/B978-0-12-401678-1.00002-6
Methods in Biomedical Informatics.
© 2014 Elsevier Inc. All rights reserved.
2.1
OBJECTIVES OF INTEGRATION
The broad objective of integration is to be able to answer questions of com-
bined data that would be otherwise very difficult and/or tedious to address
if each individual data source had to be accessed separately or in sequence.
We can look at these goals in the context of Health information exchanges
(HIE) [1], which are data repositories created geographically related to a
15
Data Integration: An Overview
Prakash M. Nadkarniand Luis N. Marenco
Yale University School of Medicine, New Haven, CT, USA
2
Chapter
CHAPTER OUTLINE
2.1 Objectives of Integration 15
2.2 Integration Approaches: Overview 17
2.2.1
Scope of this Chapter 18
2.3 Database Basics 19
2.3.1 SQL Dialects 20
2.3.2
Design for High Performance 21
2.3.3
Data Integration vs. Interoperation 21
2.4 Physical vs. Logical Integration: Pros and Cons 22
2.5 Prerequisite Subtasks 27
2.5.1 Determining Objectives 27
2.5.2
Identifying Elements: Understanding the Data Sources 28
2.5.2.1 Identifying Redundancy and Inconsistency 29
2.5.2.2 Characterizing Heterogeneity: Modeling Conflicts 31
2.5.3
Data Quality: Identifying and Fixing Errors 34
2.5.4
Documenting Data Sources and Processes: Metadata 35
2.5.4.1 Ontologies 38
2.6 Data Transformation and Restructuring 39
2.7 Integration Efforts in Biomedical Research 41
2.8 Implementation Tips 42
2.8.1
Query Tools: Caveats 42
2.8.2
The Importance of Iterative Processes 43
2.9 Conclusion: Final Warnings 44
References 44](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-24-320.jpg)
![16 CHAPTER 2 Data Integration: An Overview
consortium of stakeholders (e.g., hospitals, insurance companies, and group
practices). HIEs, whose data can be accessed when needed by any autho-
rized stakeholder, are intended to maintain an up-to-date pool of essential
information on patients within the geographical region who are associated
with any stakeholder-caregiver.
The specific goals of data integration are to:
1. Be able to look at the “Big Picture”: Organizations that carry out
identical or highly similar operations at different geographical locations
need to be able to look at consolidated summaries of structurally
identical, pooled data to know how they are performing. In other
scenarios (e.g., research consortia), different sites that function
autonomously may be generating different kinds of primary data that
focus on a broad overall problem, or with the same sources (e.g., a
common pool of patients and biospecimens): here, being able to inspect
or analyze the pooled dataset can help answer research questions. With
HIEs, pooled data facilitate epidemiological and outcomes research.
2. Identify shared elements within different sources, which can then be used
as the basis of interoperation between systems that make use of individual
sources: An example of such an operational national effort is the National
Library of Medicine’s Unified Medical Language System (UMLS) [2].
The UMLS is primarily a compendium (“meta-thesaurus”) of individual
controlled vocabularies that have achieved the status of standards for
specific biomedical applications, or within certain biomedical areas.
Different areas overlap, and therefore many elements (here, biomedical
concepts) are shared across multiple vocabularies, often with different
names. The UMLS maintains a list of biomedical concepts, a list of
synonyms (terms) for each concept and their occurrence in individual
vocabularies, along with vocabulary-specific information such as the
alphanumeric codes assigned to individual concepts. In the context of
HIEs, identification of functionally similar elements across different
systems simplifies the task of updating the patient’s medical history in
that patient’s primary medical record when the patient has been treated
somewhere else within the HIE’s geographical area.
3. Eliminate duplicated effort and errors due to non-communicating
systems: Many businesses (including hospitals), which use software
from multiple vendors, are still notorious for maintaining multiple, often
non-synchronized or out-of-date copies of essential customer/patient
data, e.g., demographics, insurance, and vital parameters such as blood
group, allergies, current ailments, current medications. In emergency
cases where a patient is seen by a caregiver/institution different from the
usual caregiver (a problem that is particularly acute in the US, where the](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-25-320.jpg)
![Native Regular Infantry.
1st Regiment, Cawnpore.
2d[39]
Regiment, Barrackpore.
3d Regiment, Phillour.
4th Regiment, Noorpore.
5th Regiment, Umballa.
6th Regiment, Allahabad.
7th Regiment, Dinapoor.
8th Regiment, Dinapoor.
9th Regiment, Allygurh.
10th Regiment, Futteghur.
11th Regiment, Allahabad.
12th Regiment, Nowgong and
Jhansi.
13th Regiment, Lucknow.
14th Regiment, Moultan.
15th Regiment, Meerut.
16th[39]
Regiment, Meean Meer.
17th Regiment, Goruckpore.
18th Regiment, Bareilly.
19th Regiment, Berhampore.
20th Regiment, Meerut.
21st Regiment, Peshawur.
22d Regiment, Fyzabad.
23d Regiment, Mhow.
24th Regiment, Peshawur.
25th Regiment, Thayet Mhow.
26th Regiment, Meean Meer.
27th Regiment, Peshawur.
28th Regiment, Shahjehanpoor.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-31-320.jpg)
![29th Regiment, Jullundur.
30th Regiment, Agra.
31st Regiment, Barrackpore.
32d Regiment, Sonthal.
33d Regiment, Hosheapore.
34th Regiment, Barrackpore.
35th Regiment, Sealkote.
36th[40]
Regiment, Jullundur.
37th[40]
Regiment, Benares.
38th[41]
Regiment, Delhi.
39th[41]
Regiment, Jelum.
40th[41]
Regiment, Dinapoor.
41st Regiment, Seetapoor.
42d Regiment, Saugor.
43d Regiment, Barrackpore.
44th Regiment, Agra.
45th Regiment, Ferozpore.
46th Regiment, Sealkote.
47th[41]
Regiment, Prome.
48th Regiment, Lucknow.
49th Regiment, Meean Meer.
50th Regiment, Nagode.
51st Regiment, Peshawur.
52d Regiment, Jubbulpoor.
53d Regiment, Cawnpore.
54th Regiment, Delhi.
55th Regiment, Nowsherah.
56th Regiment, Cawnpore.
57th Regiment, Ferozpore.
58th Regiment, Rawul Pindee.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-32-320.jpg)
![59th Regiment, Umritsir.
60th Regiment, Umballa.
61st Regiment, Jullundur.
62d Regiment, Moultan.
63d Regiment, Barrackpore.
64th Regiment, Peshawur.
65th[41]
Regiment, Dinapoor.
66th[42]
Regiment, Almora.
67th[41]
Regiment, {Etawah.
{Minpooree.
68th Regiment, Bareilly.
69th Regiment, Moultan.
70th Regiment, Barrackpore.
71st Regiment, Lucknow.
72d Regiment, Agra.
73d Regiment, Jumalpore.
74th Regiment, Cawnpore.
Irregular and Local Infantry.
1st Oude Irregular Infantry, Persadpore.
2d Oude Irregular Infantry, Secrora.
3d Oude Irregular Infantry, Gonda.
4th Oude Irregular Infantry, Lucknow.
5th Oude Irregular Infantry, Durriabad.
6th Oude Irregular Infantry, Fyzabad.
7th Oude Irregular Infantry, Lucknow.
8th Oude Irregular Infantry, Sultanpore.
9th Oude Irregular Infantry, Seetapoor.
10th Oude Irregular Infantry, Mullaong.
1st Gwalior Contingent Infantry, Gwalior.
2d Gwalior Contingent Infantry, Gwalior.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-33-320.jpg)
![Malwah Contingent, Mehidpore.
Bhopal Contingent, Sehore.
Kotah Contingent, Kurrowlee.
REGIMENTS AND STATIONS OF MADRAS ARMY—MAY 1857.
Sir Patrick Grant, Commander-in-chief.
European Cavalry.
12th Lancers (Queen’s), Madras.
Native Cavalry.
1st Madras Light Cavalry, Trichinopoly.
2d Madras Light Cavalry, Sholapore.
3d Madras Light Cavalry, Bangalore.
4th Madras Light Cavalry, Kamptee.
5th Madras Light Cavalry, Bellary.
6th Madras Light Cavalry, Jaulnah.
7th Madras Light Cavalry, Secunderabad.
8th Madras Light Cavalry, Bangalore.
European Infantry.
74th Foot (Queen’s), Madras.
84th Foot (Queen’s), Burmah.[43]
1st Europeans (East India Company’s), [Persia].
2d Europeans (East India Company’s), Burmah.
3d Europeans (East India Company’s), Secunderabad.
Native Infantry.
1st Regiment,[44]
Secunderabad.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-36-320.jpg)
![2d Regiment, Quilon.
3d Regiment, Cananore.
4th Regiment, Burmah.
5th[44]
Regiment, Berhampore.
6th Regiment, Burmah.
7th Regiment, Moulmein.
8th Regiment, Rangoon.
9th Regiment, Samulcottah.
10th Regiment, Rangoon.
11th Regiment, Cananore.
12th Regiment, Madras.
13th Regiment, Moulmein.
14th Regiment, Singapore.
15th Regiment, Burmah.
16th[44]
Regiment, Mangalore.
17th Regiment, Madras.
18th Regiment, Madras.
19th Regiment, Bangalore.
20th Regiment, French Rocks.
21st Regiment, Paulghaut.
22d Regiment, Secunderabad.
23d Regiment, Russelcondah.
24th[44]
Regiment, Secunderabad.
25th Regiment, Trichinopoly.
26th[44]
Regiment, Kamptee.
27th Regiment, Vellore.
28th Regiment, Hosungabad.
29th Regiment, Penang.
30th Regiment, Cuddapah.
31st Regiment, Vizianagram.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-37-320.jpg)
![32d Regiment, Kamptee.
33d Regiment, Kamptee.
34th Regiment, Trichinopoly.
35th Regiment, Hurryhur.
36th[44]
Regiment, Madras.
37th[45]
Regiment, Burmah.
38th[44]
Regiment, Singapore.
39th Regiment, Madras.
40th Regiment, Cuttack.
41st Regiment, Secunderabad.
42d Regiment, Secunderabad.
43d Regiment, Vizagapatam.
44th Regiment, Burmah.
45th Regiment, Rangoon.
46th Regiment, Henzana.
47th Regiment, Bellary.
48th Regiment, Moulmein.
49th[44]
Regiment, Secunderabad.
50th Regiment, Bangalore.
51st Regiment, Pallamcottah.
52d Regiment, Mercara.
Artillery, Engineers, Sappers and Miners.
Horse-artillery, 4 European
Troops.
}
Horse-artillery, 2 Native Troops. } Head-quarters:
Foot-artillery, 4 European
Battalions,
(4 Companies
each.)
}
}
St Thomas’s Mount,
Bangalore,](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-38-320.jpg)
![Foot-artillery, 1 Native Battalion.
(6 Companies.)
}
}
Kamptee, Saugor,
Secunderabad.
Engineers, Head-quarters: Fort St George.
Sappers and
Miners,
Head-quarters: Dowlaishweram.
REGIMENTS AND STATIONS OF BOMBAY ARMY—MAY 1857.
Sir Henry Somerset, Commander-in-chief.
European Cavalry.
14th Light Dragoons (Queen’s), Kirkee.
Native Regular Cavalry.
1st Lancers, Nuseerabad.
2d Light Cavalry, Rajcote.
3d Light Cavalry, [Persia.]
Native Irregular Cavalry.
1st Sinde Irregular Horse, Jacobabad.
2d Sinde Irregular Horse, Jacobabad.
Poonah Irregular Horse, [Persia.]
Gujerat Irregular Horse, Ahmedabad.
South Mahratta Irregular Horse, [Persia.]
Cutch Irregular Horse, Bhooj.
European Infantry.
64th Foot (Queen’s), [Persia.]
78th Foot (Queen’s), Poonah.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-39-320.jpg)
![86th Foot (Queen’s), Kurachee.
1st Fusiliers (East India Company’s), Kurachee.
2d Light Infantry (East India Company’s), [Persia.]
3d Light Infantry (East India Company’s), Poonah.
Native Regular Infantry.
1st Regiment,[46]
Baroda.
2d[46]
Regiment, Ahmedabad.
3d Regiment, Sholapore.
4th[47]
Regiment, [Persia.]
5th Regiment, Bombay.
6th Regiment, Poonah.
7th Regiment, Poonah.
8th Regiment, Baroda.
9th Regiment, Surat.
10th Regiment, Nuseerabad.
11th Regiment, Bombay.
12th Regiment, Deesa.
13th Regiment, Hydrabad.
14th Regiment, Kurachee.
15th Regiment, Bombay.
16th Regiment, Shikarpore.
17th Regiment, Bhooj.
18th Regiment, [Aden.]
19th Regiment, Mulligaum.
20th Regiment, [Persia]
21st Regiment, Neemuch.
22d Regiment, Satara.
23d Regiment, [Persia.]
24th Regiment, Ahmednuggur.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-40-320.jpg)
![25th Regiment, Ahmedabad.
26th Regiment, [Persia.]
27th Regiment, Kolapore.
28th Regiment, Dharwar.
29th Regiment, Belgaum.
Native Irregular Infantry.
1st Belooch Battalion, Kurachee.
2d Belooch Battalion, [Persia.]
Khandeish Bheel Corps, Dhurrungaum.
Rutnagherry Rangers, Rutnagherry.
Sawunt Waree Corps, Sawunt Waree.
Satara Local Corps, Satara.
Kolapore Infantry Corps, Kolapore.
Artillery, Engineers, Sappers and Miners.
Horse-
artillery,
1 European
Brigade.
}
(4 Troops.)[48]
} Head-quarters:
Foot-artillery, 2 European
Battalions.
} Bombay.
(4 Companies
each.)
} Ahmedabad.
Foot-artillery, 2 Native Battalions. } Ahmednuggur.
(6 Companies
each.)
}
Engineers, Head-quarters: Bombay,
Sappers and Miners, Head-quarters: Poonah
and Aden.](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-41-320.jpg)
![fearing that his proclamation had not been strong enough, he issued
another, to the effect that no new cartridges whatever should be
served out—thereby, as he hoped, putting an end to all fear
concerning objectionable lubricating substances being used; for he
was not aware how largely hypocrisy was mixed up with sincerity in
the native scruples on this point.
Anson and Barnard, when together at Umballa, had to measure well
the forces available to them. The Umballa magazines were nearly
empty of stores and ammunition; the artillery wagons were in the
depôt at Phillour; the medical officers dreaded the heat for troops to
move in such a season; and the commissariat was ill supplied with
vehicles and beasts of burden and draught. The only effectual
course was found to be, that of bringing small detachments from
many different stations; and this system was in active progress
during the week following Anson’s arrival at Umballa. On the 16th,
troops came into that place from Phillour and Subathoo. On the 17th
arrived three European regiments from the Hills,[49]
which were
shortly to be strengthened by artillery from Phillour. The prospect
was not altogether a cheering one, for two of the regiments at the
station were Bengal native troops (the 5th and 60th), on whose
fidelity only slight reliance could be placed at such a critical period.
In order that no time might he lost in forming the nucleus of a force
for Delhi, some of the troops were despatched that same night;
comprising one wing of a European regiment, a few horse, and two
guns. On successive days, other troops took their departure as
rapidly as the necessary arrangements could be made; but Anson
was greatly embarrassed by the distance between Umballa and the
station where the siege-guns were parked; he knew that a besieging
army would be of no use without those essential adjuncts; and it
was on that account that he was unable to respond to Viscount
Canning’s urgent request that he would push on rapidly towards
Delhi.
On the 23d of May, Anson sketched a plan of operations, which he
communicated to the brigadiers whose services were more
immediately at his disposal. Leaving Sir Henry Barnard in command](https://image.slidesharecdn.com/90600-250612070927-91f4d23e/85/Methods-in-Biomedical-Informatics-A-Pragmatic-Approach-1st-Edition-Neil-Sarkar-Eds-48-320.jpg)
Methods in Biomedical Informatics A Pragmatic Approach 1st Edition Neil Sarkar (Eds.)
- 1. Methods in Biomedical Informatics A Pragmatic Approach 1st Edition Neil Sarkar (Eds.) download pdf https://ebookultra.com/download/methods-in-biomedical-informatics-a- pragmatic-approach-1st-edition-neil-sarkar-eds/ Visit ebookultra.com today to download the complete set of ebook or textbook!
- 2. We have selected some products that you may be interested in Click the link to download now or visit ebookultra.com for more options!. Modal Verbs in Marlowe and Shakespeare A Semantic Pragmatic Approach 1st Edition Monika Skorasi■ska https://ebookultra.com/download/modal-verbs-in-marlowe-and- shakespeare-a-semantic-pragmatic-approach-1st-edition-monika- skorasinska/ Computational Methods in Biomedical Research 1st Edition Ravindra Khattree https://ebookultra.com/download/computational-methods-in-biomedical- research-1st-edition-ravindra-khattree/ Disease Surveillance A Public Health Informatics Approach 1st Edition Joseph S. Lombardo https://ebookultra.com/download/disease-surveillance-a-public-health- informatics-approach-1st-edition-joseph-s-lombardo/ Principles of Biomedical Informatics 2nd Edition Ashish Verma And Anchal Singh (Eds.) https://ebookultra.com/download/principles-of-biomedical- informatics-2nd-edition-ashish-verma-and-anchal-singh-eds/
- 3. Computational Methods in Biomedical Research 1st Edition Ravindra Khattree (Editor) https://ebookultra.com/download/computational-methods-in-biomedical- research-1st-edition-ravindra-khattree-editor/ Conversational Informatics An Engineering Approach 1st Edition Toyoaki Nishida https://ebookultra.com/download/conversational-informatics-an- engineering-approach-1st-edition-toyoaki-nishida/ Electronics a systems approach Fourth ed Fourth Edition Neil Storey https://ebookultra.com/download/electronics-a-systems-approach-fourth- ed-fourth-edition-neil-storey/ Ethics Research Methods and Standards in Biomedical Engineering 1st Edition Monique Frize https://ebookultra.com/download/ethics-research-methods-and-standards- in-biomedical-engineering-1st-edition-monique-frize/ Biomedical Image Understanding Methods and Applications 1st Edition Joo-Hwee Lim https://ebookultra.com/download/biomedical-image-understanding- methods-and-applications-1st-edition-joo-hwee-lim/
- 5. Methods in Biomedical Informatics A Pragmatic Approach 1st Edition Neil Sarkar (Eds.) Digital Instant Download Author(s): Neil Sarkar (Eds.) ISBN(s): 9780124016781, 0124016782 Edition: 1 File Details: PDF, 15.30 MB Year: 2014 Language: english
- 6. Methods in Biomedical Informatics A Pragmatic Approach
- 7. Methods in Biomedical Informatics A Pragmatic Approach Indra Neil Sarkar University of Vermont, Burlington, USA Edited by AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Academic Press is an lmprint of Elsevier
- 8. Academic Press is an imprint of Elsevier 32 Jamestown Road, London NW1 7BY, UK 225 Wyman Street, Waltham, MA 02451, USA 525 B Street, Suite 1800, San Diego, CA 92101-4495, USA Copyright © 2014 Elsevier Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher. Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@elsevier.com. Alternatively, visit the Science and Technology Books website at www.elsevierdirect.com/rights for further information. Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-12-401678-1 For information on all Academic Press publications visit our website at elsevierdirect.com Typeset by Scientific Publishing Services (P) Ltd., Chennai www.sps.co.in Printed and bound in United States of America 13 14 15 16 17 10 9 8 7 6 5 4 3 2 1
- 9. Contributors Gil Alterovitz Children’s Hospital Informatics Program at Harvard-MIT Division of Health Science, Boston, MA, USA. Center for Biomedical Informatics, Harvard Medical School, Boston, MA, USA. Computer Science and Artificial Intelligence Laboratory, Department of Electrical Engineering and Computer Science, MIT, Cambridge, MA, USA. Riccardo Bellazzi Dipartimento di Ingegneria Industriale e dell’Informazione, Università di Pavia, Pavia, Italy. Elizabeth S. Chen Center for Clinical and Translational Science, University of Vermont, Burlington, VT, USA Department of Medicine, Division of General Internal Medicine, University of Vermont, Burlington, VT, USA Department of Computer Science, University of Vermont, Burlington, VT, USA Hsun-Hsien Chang Children’s Hospital Informatics Program at Harvard-MIT Division of Health Science, Boston, MA, USA. Kevin Bretonnel Cohen Computational Bioscience Program, University of Colorado School of Medicine, Department of Linguistics, University of Colorado at Boulder, USA. Trevor Cohen University of Texas School of Biomedical Informatics at Houston, Houston, TX, USA. Joshua C. Denny Department of Biomedical Informatics and Medicine, Vanderbilt University, Nashville, TN, USA. Matteo Gabetta Dipartimento di Ingegneria Industriale e dell’Informazione, Università di Pavia, Pavia, Italy. Kathleen Gray Health and Biomedical Informatics Centre, Melbourne Medical School, Faculty of Medicine, Dentistry & Health Sciences and Department of Computing and Information Systems, The University of Melbourne, Melbourne, VIC, Australia. John H. Holmes University of Pennsylvania, Philadelphia, PA, USA. xv
- 10. Contributors xvi Giorgio Leonardi Dipartimento di Ingegneria Industriale e dell’Informazione, Università di Pavia, Pavia, Italy. Haiquan Li Department of Medicine, University of Illinois at Chicago, Chicago, IL, USA. Guillermo Lopez-Campos Health and Biomedical Informatics Centre, Melbourne Medical School, Faculty of Medicine, Dentistry & Health Sciences, The University of Melbourne, Melbourne, VIC, Australia. Yves A. Lussier Department of Medicine, University of Illinois at Chicago, Chicago, IL, USA. Department of Bioengineering, University of Illinois at Chicago, Chicago, IL, USA. Cancer Center, University of Illinois, Chicago, IL, USA. Luis N. Marenco Yale University School of Medicine, New Haven, CT, USA. Fernando Martin-Sanchez Health and Biomedical Informatics Centre, Melbourne Medical School, Faculty of Medicine, Dentistry & Health Sciences, The University of Melbourne, Melbourne, VIC, Australia. Jason H. Moore Dartmouth College Department of Genetics, Hanover, NH, USA. Mark A. Musen Stanford Center for Biomedical Informatics Research, Stanford University, Stanford, CA, USA. Prakash M. Nadkarni Yale University School of Medicine, New Haven, CT, USA. Indra Neil Sarkar Center for Clinical and Translational Science, University of Vermont, Burlington, VT, USA. Ryan J. Urbanowicz Dartmouth College Department of Genetics, Hanover, NH, USA. Dominic Widdows Microsoft Bing, Serendipity, Palo Alto, CA, USA. Hua Xu School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, TX, USA.
- 11. Methods in Biomedical Informatics. © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/B978-0-12-401678-1.00001-4 1 Chapter 1.1 BIOMEDICAL INFORMATICS AND ITS APPLICATIONS Biomedicine is a complex ecosystem of foundational and applied disciplines. Generally, researchers and practitioners in biomedicine are specialized in a particular area of emphasis that can be described as either: (1) “bench”— aiming to understand the underpinning principles of dysfunction leading to compromised health status; (2) “bedside”—aiming to develop preventative approaches for the maintenance of normal health status or addressing abnor- mal health status through clinical intervention and quantifying the effect on individuals; or (3) “community”—aiming to quantify the effect of preventa- tive measures or treatment regimens across defined population groups and developing systematic approaches to promote effective practices. There has been increased discussion in recent years about these areas of emphasis and the need for strategies to promote synergies between them [1–3]. This book is focused on describing methodologies that underpin the disci- pline of Biomedical Informatics, which is defined as a field “that studies and pursues the effective uses of biomedical data, information, and knowledge for scientific inquiry, problem solving and decision making, motivated by efforts to improve human health [4].” Biomedical informatics has evolved into a discipline that encompasses the bench, bedside, and community areas of biomedicine and represents a synergy between foundational and applied aspects of biomedical science [4]. The foundational aspects of biomedi- cal informatics are derived from a wide array of fundamental disciplines, 1 CHAPTER OUTLINE 1.1 Biomedical Informatics and its Applications 1 1.2 The Scientific Method 4 1.3 Data, Information, Knowledge, and Wisdom 6 1.4 Overview of Chapters 9 1.5 Expectations and Challenge to the Reader 11 References 12 Introduction Indra Neil Sarkar Center for Clinical and Translational Science, University of Vermont, Burlington, VT, USA
- 12. 2 CHAPTER 1 Introduction including those from formal science (e.g., logic, mathematics, and statistics) and basic science (e.g., physics, chemistry, biology, psychology, and eco- nomics). Additional applied disciplines also contribute key foundational elements to biomedical informatics (e.g., medical science, computer sci- ence, engineering, and epidemiology). From the foundational sciences that contribute to biomedical informatics, methodologies are developed and applied to specific biomedical areas. For example, biomedical informatics methods may be applied to spe- cific domain areas, like: biology (termed “bioinformatics”), medicine and nursing (termed “clinical informatics”), and public health (termed “pub- lic health informatics”). In contemporary biomedical informatics par- lance, “health informatics” is often used as an umbrella term for clinical informatics and public health informatics. There are many other areas of informatics that traverse domain boundaries (e.g., “imaging informatics” involves the application of biomedical informatics approaches to areas that span biological and clinical domains) or are aligned with particular spe- cialties (e.g., “pathology informatics”). A comprehensive survey of bio- medical informatics and its application areas can be found in Biomedical Informatics: Computer Applications in Health Care (edited by Shortliffe and Cimino) [5]. Biomedical informatics methods may also be applied to areas that focus on the translation of innovations from one part of the biomedicine spectrum to another [6]. In recent years, this area of emphasis has collectively been referred to as “translational sciences.” Working alongside translational med- icine, the translational sciences aim to enable greater synergy between basic and applied sciences [7,8]. Within the context of biomedical informatics, “translational informatics” is the use of informatics approaches to support translational science and medicine [9]. Translational informatics is divided into two major sub-specialties: (1) “translational bioinformatics”—which is focused on the development of informatics approaches that facilitate the translation of basic science findings into ones that may be utilized in clinical contexts [10,11]; and (2) “clinical research informatics”—which is focused on the development of informatics approaches for evaluating and gener- alizing clinical innovations [12]. Translational bioinformatics and clinical research informatics work in tandem to increase the likelihood of basic sci- ence innovations being put into clinical practice and improving the health of communities—thus directly addressing the translational bottlenecks, such as those that are often seen at the transition of knowledge from bench to bedside or bedside to community. Figure 1.1 provides a graphical overview biomedical informatics and its sub-disciplines that span the full breath of biomedicine.
- 13. 3 1.1 BiomedicalInformaticsanditsApplications The expansiveness of biomedical informatics in many ways requires a poly- mathic approach to developing approaches for the betterment of health. Formal training in biomedical informatics thus requires a unique combina- tion of formal, basic, and applied science [4]. It therefore is not uncommon for one to first attain formal, graduate-level scholarly or professional train- ing in at least one area of science before embarking on additional training in biomedical informatics. In addition to these cross-trained individuals, there are also a growing number of formally trained biomedical informaticians whose entire graduate-level education is done in biomedical informatics. In most cases, biomedical informaticians choose at least one application area of specialization (e.g., bioinformatics, clinical informatics, or public health informatics). Regardless of the path to becoming a biomedical infor- matician, the approaches used to address biomedical problems are built on a common set of methodologies [4]. It must be noted that the highly special- ized training required for success in biomedical informatics has resulted in a significant shortage of biomedical informaticians across the entire spectrum of biomedicine. To address this challenge, there are an increasingly grow- ing number of formal training opportunities that strive to help provide the biomedical enterprise with biomedical informaticians [13]. In contrast to multi- or inter-disciplinary disciplines, where foundational ele- ments are respectively combined in additive or interactive ways, biomedical informatics is a trans-disciplinary discipline, where foundational elements are holistically combined in ways that result in the emergence of entirely new concepts [14–16]. To this end, biomedical informatics is a unique dis- cipline in that it brings together an array of diverse expertise and experi- ence, but remains with the singular purpose to develop methods to improve the process of health maintenance and treatment of deviations from normal Bench Bedside Bioinformatics Clinical Informatics Community Public Health Informatics Biomedical Informatics Health Informatics Translational Bioinformatics Clinical Research Informatics Translational Sciences Domain Sciences Translational Informatics n FIGURE 1.1 Overview of Biomedical Informatics. Major application areas of biomedical informat- ics are shown, aligned according to their general emphasis to develop solutions for bench, bedside, or community stakeholders. The major application areas are also segregated according to their domain or translational science centricity. Note that not all application areas are shown.
- 14. 4 CHAPTER 1 Introduction health. The development and use of these methodologies can be organized according to the Scientific Method (described in Section 1.2), with the goal of transforming biomedical data into information that leads to actionable knowledge and therefore leading to wisdom (collectively referred to as the DIKW framework, described in Section 1.3). This book provides an intro- duction to a number of key methodologies used in biomedical informatics that hone the Scientific Method in the context of the DIKW framework (an overview of the chapters is provided in Section 1.4, along with expectations in Section 1.5). 1.2 THE SCIENTIFIC METHOD Like all scientific disciplines, biomedical informatics is strongly grounded in the Scientific Method. The Scientific Method can be traced back to Artistotle, who introduced the principles of logic coming in two forms [17]: (1) inductive—which makes postulations based on observation of universal concepts; and (2) deductive—which makes postulations based on relation- ships between already accepted universal concepts (called “syllogisms”). It was not, however, until the Arab polymath al-Hasan ibn al-Haytham (often referred to in the Western world as “Alhazen” or “Alhacen”) described the principles of optics in a systematic manner that the contemporary Scientific Method became a formally described process [18]. Ibn al-Haytham’s Book of Optics was one of the main sources used by the Englishman Roger Bacon (not to be confused with Francis Bacon, who described an alternative induc- tive methodology referred to as the “Baconian Method,” which in many ways rejected the hypothesis-driven Scientific Method [19]) to formally describe the Scientific Method to the Western world [20]. The Scientific Method consists of five major activities: (1) Question for- mulation; (2) Hypothesis Generation; (3) Prediction; (4) Testing; and (5) Analysis. The first activity of question formulation aims to identify a query of interest relative to a specific observation (e.g., “will this treatment regi- men cure the patient of this illness?”). Question formulation can involve the consultation of existing knowledge sources or other experts for determining the validity of the question. Question formulation, in many ways, is the most difficult step of the Scientific Method; however, it is also the most crucial step because all the consequent steps are dependent on a well-formed ques- tion. Once a question is developed, the next stage of the Scientific Method strives to add focus by postulating a specific hypothesis (e.g., “this treatment will treat the patient of their illness.”). The generation of a hypothesis is often done in two parts, which together make the hypothesis testable: (1) defining the null hypothesis (H0)—a con- templation of the hypothesis in a statistical framework that presents the
- 15. 5 1.2 TheScientificMethod default conclusion (e.g., “this treatment regimen does not cure the patient of their illness compared to a placebo.”); and (2) defining the alternative hypothesis (H1)—a contemplation of the hypothesis in a statistical frame- work that presents the desired outcome (e.g., “this treatment, when com- pared to a placebo, cures the patient of their illness”). An important feature of a well-formed hypothesis is that it is falsifiable. Falsifiability is defined as the ability to identify potential solutions that could disprove a stated hypoth- esis (e.g., “the patient is cured by a placebo.”). Thus, the testability of a hypothesis is inherently a feature of its falsifiability [21]. After it is deemed that a hypothesis is testable, the next stage in the Scientific Method is to propose some predictions that help determine the plausibility of the hypothesis relative to alternatives, including coincidence (e.g., “this treatment cures the patient of their illness compared to doing nothing”). The predictions are then tested through the gathering of evidence, which support or refute the hypothesis of interest (e.g., “for a sample of chosen patients who have the same illness, measure the effect of the treatment versus a pla- cebo versus nothing”). As noted earlier, an important feature of developed tests is that they aim to address the falsifiability of the hypothesis. The results of the tests are then analyzed to determine if the hypothesis was indeed proved true (rejection of the null hypothesis and thus acceptance of the alternative hypothesis). This is commonly done using a statistical com- parison test such as one that can be derived from a confusion matrix [which delineates verifiable (“true positive” and “true negative”) and unexpected (“false positive” and “false negative”) results based on previous knowl- edge]. Common comparison tests include the Pearson’s chi-square [22] and Fisher’s exact test [23]. The final outcome of the analysis is a statement relative to the original question (e.g., “yes, this treatment regimen will cure patients of this illness.”). The Scientific Method does not necessarily conclude at the completion of the analysis step; the analysis of one hypothesis may lead to additional questions that can consequently be examined through an additional itera- tion of the Scientific Method. Indeed, the Scientific Method as classically implemented can be perceived as an infinite process. Within the context of biomedicine, a modern interpretation of the Scientific Method is often used, termed the “hypothetico-deductive approach.” The hypothetico-deductive approach is a cornerstone in clinical education and practice [24–26]: (1) gather data about a patient; (2) develop questions that lead to hypotheses for the patient’s state; (3) propose predictions based on the suggested hypotheses that explain the patient’s state; and (4) test the hypotheses through attempts at falsifying them to explain the patient’s state. It is important to once again
- 16. 6 CHAPTER 1 Introduction underscore that clinical inquiry into patient status is not a verification of a particular condition; evidence to support a proposed set of hypotheses is done through a series of falsification tests (in clinical parlance, these are often termed “rule-in” or “rule-out” tests). As in many clinical scenarios, it must be acknowledged that the absolute true diagnosis for a given patient may not be known or even knowable (e.g., to determine whether a patient actually has Alzheimer disease, the most definitive method of detection is a neuropathological analysis done post-mortem [27]); however, the develop- ment of hypotheses and tests to prove or refute them enables a clinician to get closer to the true diagnosis. 1.3 DATA, INFORMATION, KNOWLEDGE, AND WISDOM Data are composed of individual datum points that may originate from a plethora of sources (it is for this reason that in scientific writing the term “data” should be treated as a plural noun; the singular form is “datum”). Simply put, data are the raw substrate that can be transformed through for- mal science methods to acquire meaning. Data may come in many forms, from any part of the biomedical spectrum. Data are the essential build- ing blocks wherefrom the Scientific Method begins and may lead to the gathering of additional data in the testing of hypotheses. In biomedicine, data are generated from a wide range of sources—as artifacts of digital systems (e.g., as might be generated from automated laboratory systems) or as recorded events from human-human interactions (e.g., as might be generated from a clinician-patient interview). Data are an essential com- ponent of modern science, and can acquire different meanings depending on their interpretation. These interpretations are dependent on the use of appropriate formal science techniques—most often a combination of logic and statistics. It is important for one to be aware of the assumptions made in both the logic (e.g., the definition of true versus false) and statistics (e.g., the data distributed in a Gaussian manner). An additional layer of complexity is that data can, and often do, suffer from unavoidable inconsistencies, which may be an artifact of either the generation, collection, or interpretation of data. Even with all these issues, data form the basis for all derived interpretations and thus form the foundation for the biomedical sciences. It is the study and improvement of formal science techniques that also form the basis of biomedical informatics methods. As data are transformed into information, which is the result of the appli- cation of formal science techniques, hypotheses may be generated about
- 17. 7 1.3 Data,Information,Knowledge,andWisdom the meaning of the observed data that form the basis for basic science. The basic sciences, which have a strong basis in the Scientific Method (built around the generation, testing, and validation of hypotheses), impart interpretations of data based on a specific domain focus. For example, physics is focused on the interpretation of data to understand the physical world in terms of matter and its motion; chemistry is focused on analyzing data to understand the composition of matter; biology is focused on understanding extinct and extant organisms; and economics is focused on the understanding of data associated with goods or services that form the basis for inter-personal relationships. The basic sciences form the foundation for biomedicine, and provide the insights about the underlying cause of dysfunction and its detection as well as provide the tools for tracking treatment outcomes from both clinical and economical perspectives. As information about health is collected and analyzed, it can be coalesced into reusable constructs (e.g., a particular treatment regimen that has been shown to improve health). These constructs form the basis of knowledge, which require a systematic understanding and use of data that have been transformed into interpretable information. Knowledge can be used to guide decisions or subsequent analyses—this type of knowledge is referred to as “actionable knowledge [28].” Actionable knowledge is of the most utility in biomedicine when data can be used in a way to guide clinical decisions in a manner to positively affect patient health outcomes. The study and evalua- tion of knowledge leads to wisdom, which promotes the development of best practices and guidance for how to interpret future encounters with biomedi- cal data. The applied sciences provide the scaffolding to systematically transform information from biomedical data into knowledge and retain it as wisdom that can be used to guide disease diagnoses, treatment regimens, or out- comes analyses. For example, medical science approaches can be used to consistently interpret a laboratory result in combination with a collection of signs and symptoms to determine the health status of a patient; engineering approaches can be used to develop a process to provide consistent levels of care to patients as they interact with the health care system; and epidemi- ology approaches can be used to analyze the effect of vaccinations across susceptible populations. Data, information, knowledge, and wisdom and their relationships to each other are collectively referred to as the Data-Information-Knowledge- Wisdom (DIKW) framework, and the most common interpretations are cred- ited to Ackoff [29, 30]. The DIKW framework (where data are transformed
- 18. 8 CHAPTER 1 Introduction into wisdom) can certainly be applied to each individual scientific disci- pline; however, as noted earlier, what sets biomedical informatics apart from other disciplines is that its scope is trans-disciplinary. The DIKW frame- work can be applied in specific formal, basic, or applied science contexts; however, in the context of biomedical informatics, the DIKW framework is used to bridge formal, basic, and applied sciences toward a single purpose— to improve the diagnosis, care, and treatment of illness. Furthermore, the DIKW framework in biomedical informatics formalizes the implementa- tion of the Scientific Method towards the discovery and implementation of biomedical innovations. Like the Scientific Method, the DIKW framework does not necessarily conclude at the establishment of wisdom; the wisdom gained from one set of data can be used as data for a subsequent study. The relationships between the sciences, the Scientific Method, and the DIKW framework are graphically depicted in Figure 1.2. The DIKW framework can be used to formally organize, study, and innovate at different levels of inquiry. Put into the context of biomedicine, the DIKW framework is an essential process that transverses the aforementioned major areas of bench, bedside, and community. The results of transforming data into wisdom in one area may very well lead to data for another area. Of course, the boundaries between the suggested areas of biomedicine are not always apparent. Recently, there has been a concerted effort in the biomedi- cal research enterprise to acknowledge that much of biomedicine innovation does suffer from a “siloed” approach that force the bench, bedside, and com- munity researchers into disjointed endeavors [1–3]. From the earliest days of biomedical informatics, it was proposed that con- cepts and approaches, such as those that could be implemented in computers, could be used to better integrate the bench, bedside, and community areas of research and practice [31–33]. Biomedical informaticians thus necessarily Data Information Knowledge Wisdom Scientific Method Question Formulation Hypothesis Generation Prediction Testing Analysis e.g., Mathematics, Logic, Statistics Formal Sciences Basic Sciences e.g., Physics, Chemistry, Biology, Economics Applied Sciences e.g. Medical Science, Engineering, Epidemiology n FIGURE 1.2 DIKW Framework. The Data-Information-Knowledge-Wisdom (DIKW) framework is depicted as a process that unifies the formal, basic, and applied sciences. The graphic also shows how the major steps of the Scientific Method can be aligned with the transformation of data into wisdom.
- 19. 9 1.4 OverviewofChapters work together as teams of formal, basic, and applied scientists to develop solu- tions that aim to address specific bench, bedside, or community challenges. Whilst a given approach may be originally designed for one particular area, biomedical informatics espouses that it may be generalized to another area. The realization of the DIKW framework in the practice of biomedical infor- matics is thus intertwined in a manner that suggests a holistic approach that uniquely unifies the many facets of biomedicine (as depicted in Figure 1.3). 1.4 OVERVIEW OF CHAPTERS The main goal of this book is to present biomedical informatics methods that are used to focus the Scientific Method to transform data into wisdom, along the aforementioned DIKW framework. Where possible, practical examples are provided that aim to help the reader appreciate the methodological concepts within “real world” bench, bedside, or community scenarios. It is impossible to provide complete coverage of the entire range of biomedi- cal informatics methods. Therefore, this book aims to provide a foundation for many of the commonly used and discussed approaches. Similarly, it is impossible to fully describe all the nuances of a given methodology across all possible biomedical scenarios in a single book chapter. The chapters of this book thus focus on presenting key features of a given area of biomedi- cal informatics methodology with emphasis on a chosen set of biomedical contexts. All of the authors are established leaders in the development and application of biomedical informatics methods, and present examples from their own work and experience. The overall order of chapters in this book aims to present methodologies according to the DIKW framework, and a given chapter may span multiple facets of the framework. As described earlier, the substrate wherefrom biomedical informatics inno- vations emerge to address challenges in biomedicine is composed of data. It is not uncommon for one to bring together multiple streams of data to develop transformative approaches further along the DIKW framework. Chapter 2 (by Prakash M. Nadkarni and Luis N. Marenco) thus focuses on Data Information Knowledge Wisdom Data Information Knowledge Wisdom Data Information Knowledge Wisdom Bench Bedside Community n FIGURE 1.3 Iterative DIKW Process to Translation. The DIKW framework is shown as an iterative process where wisdom that is derived from bench, bedside, or community-based data can be used as source data for generation of additional testable hypotheses.
- 20. 10 CHAPTER 1 Introduction the methodologies associated with data integration. Particular attention is given to comparing competing approaches for integrating data from dispa- rate sources, while accounting for political and resource realities. Chapter 3 (by Mark A. Musen) then shifts the attention to how one might represent knowledge that can be attributed to gathered data. Within the context of the DIKW framework, this chapter highlights important methodological con- siderations and challenges with representing the meaning and preserving the knowledge associated with data.As continual improvements are seen in data generation technologies, such as next generation molecular sequencing, there is an increased need to harness biomedical informatics methodologies to identify potential testable hypotheses. To this end, Chapter 4 (by Yves A. Lussier and Haiquan Li) explores the challenges with generating hypotheses from heterogeneous data sets that span the spectrum of biomedicine. It is essential to acknowledge that a significant volume of biomedical knowl- edge is not readily searchable or available for computational analyses. The next two chapters thus aim to introduce the reader to key methods associated with retrieval of knowledge from traditionally text based sources. The first of these, Chapter 5 (by Trevor Cohen and Dominic Widdows), presents contem- porary biomedical informatics approaches that utilize geometric techniques to explore or analyze multi-faceted biomedical knowledge. Chapter 6 (by Kevin B. Cohen) provides an overview of natural language processing, which continues to mature within the realm of biomedical informatics for extracting potentially usable information from a range of sources across biomedicine. As data are gathered from a plethora of sources and potentially represented as knowledge that can be of utility for future use, one must consider how one performs the transformation from data into information and knowledge such that it might enter into the stage of wisdom. Chapter 7 (by John H. Holmes) provides a foundational overview of data mining techniques, which are used to realize the DIKW framework. The next two chapters then present specific techniques that are used in biomedical informatics to impute information and knowledge from biomedical data. Chapter 8 (by Hsun-Hsien Chang and Gil Alterovitz) presents Bayesian methods that represent a major foundational category of techniques used in biomedical informatics to impute knowl- edge from biomedical data. Chapter 9 (by Ryan J. Urbanowicz and Jason H. Moore) then introduces learning classifier systems, which are increasingly becoming essential to decipher complex phenomena from potentially unde- cipherable data that can be perceived as knowledge. As biomedical informatics solutions are developed, largely through the har- nessing of formal and basic science techniques, their biomedical utility can only be realized through the implementation and contextualization through
- 21. 11 1.5 ExpectationsandChallengetotheReader applied science. The next set of chapters aim to provide examples of method- ologies that harness the applied aspects of biomedical informatics. The first of these chapters, Chapter 10 (by Riccardo Bellazzi, Matteo Gabetta, and Giorgio Leonardi), describes fundamental engineering principles that are germane to the design, development, and ultimate implementation of bio- medical informatics innovations. Chapter 11 (by Fernando Martin-Sanchez, Guillermo Lopez-Campos, and Kathleen Gray) follows by exploring the biomedical informatics landscape associated with personalized medicine and participatory health, which reflects a holistic biomedical revolution that seamlessly integrates classical biomedical data with patient centered data to result in a new cadre of biomedical knowledge. Chapter 12 (by Joshua C. Denny and Hua Xu) then focuses on the development of personalized care regimens that harness increasingly digitally available genomic and health data that can be used to develop informed clinical decisions. Chapter 13 provides a concluding perspective of biomedical informatics and its continued relevance in the emerging “Big Data” era. The chapters of the main book are followed by four mini-primers (Appendices A–D by Elizabeth S. Chen) that aim to provide hands-on experience with basic technical skills that are often used in the implementation of biomedical informatics methods: Unix (Appendix A); Ruby (Appendix B); Databases (Appendix C); and Web Services (Appendix D). 1.5 EXPECTATIONS AND CHALLENGE TO THE READER As with any multi-authored book, there will be varying styles in presenting content. Each of the authors was charged with taking what are tradition- ally complex and difficult concepts in biomedical informatics and present- ing them in a manner that is accessible but still of utility in the context of contemporary biomedicine. To this end, each chapter should be approached with the aim to understand the principles of the concepts described within the context of the examples from within a given chapter. The reader should then aim to address three questions: 1. What are the key aspects of the methodological concepts described? 2. How can the methodological concepts be applied to my area of interest (e.g., bioinformatics, clinical informatics, or public health informatics)? 3. What are potential advantages/disadvantages of the methods presented? The reader should then aim to identify additional peer-reviewed literature (starting with other articles written by the chapter authors) that further describes the methodological aspects of the techniques presented within a given chapter.
- 22. 12 CHAPTER 1 Introduction For this book to be used as an effective educational instrument or as an intro- duction to the broad range of methodologies that are used in biomedical infor- matics, it must be used as a starting point and not as a comprehensive reference for a given methodology. Many of the chapter authors have included references that may be consulted for additional details. Whilst a book of this nature will never be comprehensive (or even complete in topics covered), it is expected that it will provide a foundation for the methodological principles of biomedi- cal informatics that can be applied across the spectrum of biomedicine. REFERENCES [1] Keramaris NC, Kanakaris NK, Tzioupis C, Kontakis G, Giannoudis PV. Translational research: from benchside to bedside. Injury 2008;39(6):643–50. [2] Woolf SH. The meaning of translational research and why it matters. J Am Med Assoc 2008;299(2):211–3. [3] Westfall JM, Mold J, Fagnan L. Practice-based research–“Blue Highways” on the NIH roadmap. J Am Med Assoc 2007;297(4):403–6. [4] Kulikowski CA, Shortliffe EH, Currie LM, Elkin PL, Hunter LE, Johnson TR, et al. AMIA board white paper: definition of biomedical informatics and specification of core competencies for graduate education in the discipline. J Am Med Inform Assoc 2012;19(6):931–8. [5] Shortliffe EH, Cimino JJ, editors. Biomedical informatics: computer applications in health care and biomedicine. 4th ed. New York: Springer; 2013. [6] Sarkar IN. Biomedical informatics and translational medicine. J Transl Med 2010;8:22. PubMed PMID: 20187952. [7] Lean ME, Mann JI, Hoek JA, Elliot RM, Schofield G. Translational research. BMJ 2008;337:a863. [8] Wehling M. Principles of translational science in medicine: from bench to bedside. Cambridge, New York: Cambridge University Press; 2010. xxii, 382 p., 24 p. of plates p. [9] Payne PR, Embi PJ, Sen CK. Translational informatics: enabling high-throughput research paradigms. Physiol Genomics 2009;39(3):131–40. [10] Altman RB. Translational bioinformatics: linking the molecular world to the clinical world. Clin Pharmacol Ther 2012;91(6):994–1000. [11] Sarkar IN, Butte AJ, LussierYA, Tarczy-Hornoch P, Ohno-Machado L. Translational bioinformatics: linking knowledge across biological and clinical realms. J Am Med Inform Assoc 2011;18(4):354–7. [12] Embi PJ, Payne PR. Clinical research informatics: challenges, opportunities and defi- nition for an emerging domain. J Am Med Inform Assoc 2009;16(3):316–27. [13] Shortliffe EH. The future of biomedical informatics: a perspective from academia. Stud Health Technol Inform 2012;180:19–24. [14] Choi BC, Pak AW. Multidisciplinarity, interdisciplinarity and transdisciplinarity in health research, services, education and policy: 1. Definitions, objectives, and evidence of effectiveness. Clin Invest Med. Medecine clinique et experimentale 2006;29(6):351–64. [15] Choi BC, Pak AW. Multidisciplinarity, interdisciplinarity, and transdisciplinarity in health research, services, education and policy: 2. Promotors, barriers, and
- 23. 13 References strategies of enhancement. Clin Invest Med. Medecine clinique et experimentale 2007;30(6):E224–32. [16] Choi BC, Pak AW. Multidisciplinarity, interdisciplinarity, and transdisciplinarity in health research, services, education and policy: 3. Discipline, inter-discipline dis- tance, and selection of discipline. Clin Invest Med. Medecine clinique et experimen- tale 2008;31(1):E41–8. [17] Barnes J. The Cambridge companion to Aristotle. Cambridge, NewYork: Cambridge University Press; 1995. xxv, p. 404. [18] Omar SB. Ibn al-Haytham’s optics: a study of the origins of experimental science. Minneapolis: Bibliotheca Islamica; 1977. p. 168. [19] Bacon F, Jardine L, Silverthorne M. The new organon. Cambridge U.K., New York: Cambridge University Press; 2000. xxxv, p. 252. [20] Hackett J. Roger Bacon and the sciences: commemorative essays. Leiden, NewYork: Brill; 1997. x, p. 439. [21] Popper KR. The logic of scientific discovery. London, New York: Routledge; 1992. p. 479. [22] Pearson K. On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. Philos Mag Ser 5 1900;50(302):157–75. [23] Fisher RA. On the interpretation of χ2 from contingency tables, and the calculation of P. J Roy Stat Soc 1922;85(1):87–94. [24] Barrows HS, Tamblyn RM. Problem-based learning: an approach to medical educa- tion. New York: Springer Pub. Co.; 1980. xvii, p. 206. [25] Mandin H, Jones A, Woloschuk W, Harasym P. Helping students learn to think like experts when solving clinical problems. Acad Med: J Assoc Am Med Coll 1997;72(3):173–9. [26] Connelly DP, Johnson PE. The medical problem solving process. Hum Pathol 1980;11(5):412–9. [27] Beach TG, Monsell SE, Phillips LE, Kukull W. Accuracy of the clinical diagno- sis of Alzheimer disease at National Institute on Aging Alzheimer Disease Centers, 2005–2010. J Neuropathol Exp Neurol 2012;71(4):266–73. [28] Cao L. Domain driven data mining. New York, London: Springer; 2010. xvi, p. 248 [29] Ackoff RL. From data to wisdom. J Appl Syst Anal 1989;16:3–9. [30] Rowley J. The wisdom hierarchy: representations of the DIKW hierarchy. J Inf Sci 2007;33(2):163–80. [31] Ledley RS, Lusted LB. Reasoning foundations of medical diagnosis; symbolic logic, probability, and value theory aid our understanding of how physicians reason. Science 1959;130(3366):9–21. [32] Lusted LB, Ledley RS. Mathematical models in medical diagnosis. J Med Educ 1960;35:214–22. [33] Blois MS. Information holds medicine together. MD Comput: Comput Med Practice 1987;4(5):42–6.
- 24. http://dx.doi.org/10.1016/B978-0-12-401678-1.00002-6 Methods in Biomedical Informatics. © 2014 Elsevier Inc. All rights reserved. 2.1 OBJECTIVES OF INTEGRATION The broad objective of integration is to be able to answer questions of com- bined data that would be otherwise very difficult and/or tedious to address if each individual data source had to be accessed separately or in sequence. We can look at these goals in the context of Health information exchanges (HIE) [1], which are data repositories created geographically related to a 15 Data Integration: An Overview Prakash M. Nadkarniand Luis N. Marenco Yale University School of Medicine, New Haven, CT, USA 2 Chapter CHAPTER OUTLINE 2.1 Objectives of Integration 15 2.2 Integration Approaches: Overview 17 2.2.1 Scope of this Chapter 18 2.3 Database Basics 19 2.3.1 SQL Dialects 20 2.3.2 Design for High Performance 21 2.3.3 Data Integration vs. Interoperation 21 2.4 Physical vs. Logical Integration: Pros and Cons 22 2.5 Prerequisite Subtasks 27 2.5.1 Determining Objectives 27 2.5.2 Identifying Elements: Understanding the Data Sources 28 2.5.2.1 Identifying Redundancy and Inconsistency 29 2.5.2.2 Characterizing Heterogeneity: Modeling Conflicts 31 2.5.3 Data Quality: Identifying and Fixing Errors 34 2.5.4 Documenting Data Sources and Processes: Metadata 35 2.5.4.1 Ontologies 38 2.6 Data Transformation and Restructuring 39 2.7 Integration Efforts in Biomedical Research 41 2.8 Implementation Tips 42 2.8.1 Query Tools: Caveats 42 2.8.2 The Importance of Iterative Processes 43 2.9 Conclusion: Final Warnings 44 References 44
- 25. 16 CHAPTER 2 Data Integration: An Overview consortium of stakeholders (e.g., hospitals, insurance companies, and group practices). HIEs, whose data can be accessed when needed by any autho- rized stakeholder, are intended to maintain an up-to-date pool of essential information on patients within the geographical region who are associated with any stakeholder-caregiver. The specific goals of data integration are to: 1. Be able to look at the “Big Picture”: Organizations that carry out identical or highly similar operations at different geographical locations need to be able to look at consolidated summaries of structurally identical, pooled data to know how they are performing. In other scenarios (e.g., research consortia), different sites that function autonomously may be generating different kinds of primary data that focus on a broad overall problem, or with the same sources (e.g., a common pool of patients and biospecimens): here, being able to inspect or analyze the pooled dataset can help answer research questions. With HIEs, pooled data facilitate epidemiological and outcomes research. 2. Identify shared elements within different sources, which can then be used as the basis of interoperation between systems that make use of individual sources: An example of such an operational national effort is the National Library of Medicine’s Unified Medical Language System (UMLS) [2]. The UMLS is primarily a compendium (“meta-thesaurus”) of individual controlled vocabularies that have achieved the status of standards for specific biomedical applications, or within certain biomedical areas. Different areas overlap, and therefore many elements (here, biomedical concepts) are shared across multiple vocabularies, often with different names. The UMLS maintains a list of biomedical concepts, a list of synonyms (terms) for each concept and their occurrence in individual vocabularies, along with vocabulary-specific information such as the alphanumeric codes assigned to individual concepts. In the context of HIEs, identification of functionally similar elements across different systems simplifies the task of updating the patient’s medical history in that patient’s primary medical record when the patient has been treated somewhere else within the HIE’s geographical area. 3. Eliminate duplicated effort and errors due to non-communicating systems: Many businesses (including hospitals), which use software from multiple vendors, are still notorious for maintaining multiple, often non-synchronized or out-of-date copies of essential customer/patient data, e.g., demographics, insurance, and vital parameters such as blood group, allergies, current ailments, current medications. In emergency cases where a patient is seen by a caregiver/institution different from the usual caregiver (a problem that is particularly acute in the US, where the
- 26. Random documents with unrelated content Scribd suggests to you:
- 27. system in recognising the just claims of the officers more immediately in its service, and of the widows and children of those who fell during the mutiny—a system based on the established emoluments and pensions of all in the Company’s service. It will thus be seen that the news of the Indian Revolt, when it reached London by successive mails, led to a remarkable and important series of suggestions and plans—intended either to strengthen the hands of the executive in dealing with the mutineers, or to succour those who had been plunged into want by the crimes of which those mutineers were the chief perpetrators.
- 28. Note. At the end of the last chapter a table was given of the number of troops, European and native, in all the military divisions of India, on the day when the mutiny commenced at Meerut. It will be convenient to present here a second tabulation on a wholly different basis—giving the designations of the regiments instead of the numbers of men, and naming the stations instead of the divisions in which they were cantoned or barracked. This will be useful for purposes of reference, in relation to the gradual annihilation of the Bengal Hindustani army. The former table applied to the 10th of May 1857; the present will apply to a date as near this as the East India Register will permit—namely, the 6th of May; while the royal troops in India will be named according to the Army List for the 1st of May—a sufficiently near approximation for the present purpose. A few possible sources of error may usefully be pointed out. 1. Some or other of the India regiments were at all times moving from station to station; and these movements may in a few cases render it doubtful whether a particular corps had or had not left a particular station on the day named. 2. The station named is that of the head-quarters and the bulk of the regiment: detachments may have been at other places. 3. The Persian and Chinese wars disturbed the distribution of troops belonging to the respective presidencies. 4. The disarming and disbanding at Barrackpore and Berhampore are not taken into account; for they were not known in London at the time of compiling the official list. 5. The Army List, giving an enumeration of royal regiments in India, did not always note correctly the actual stations at a particular time. These sources of error, however, will not be considerable in amount. REGIMENTS AND STATIONS OF BENGAL ARMY—MAY 1857. General Anson, Commander-in-chief. European Cavalry. 6th Carabiniers (Queen’s), Meerut. 9th Lancers (Queen’s), Umballa. Native Regular Cavalry. 1st Regiment, Mhow.
- 29. 2d Regiment, Cawnpore. 3d Regiment, Meerut. 4th Regiment, Umballa. 5th Regiment, Peshawur. 6th Regiment, Nowgong. 7th Regiment, Lucknow. 8th Regiment, Lahore. 9th Regiment, Sealkote. 10th Regiment, Ferozpore. Irregular and Local Cavalry. 1st Bengal Ir. C., Jelum. 2d Bengal Ir. C., Goordaspore. 3d Bengal Ir. C., Jhansi. 4th Bengal Ir. C., Hansi. 5th Bengal Ir. C., Sonthal. 6th Bengal Ir. C., Moultan. 7th Bengal Ir. C., Peshawur. 8th Bengal Ir. C., Sultanpore. 9th Bengal Ir. C., Hosheapore. 10th Bengal Ir. C., Goordaspore. 11th Bengal Ir. C., Berhampore. 12th Bengal Ir. C., Segowlie. 13th Bengal Ir. C., Bareilly, 14th Bengal Ir. C., Jhansi. 15th Bengal Ir. C., Oude. 16th Bengal Ir. C., Rawul Pindee. 17th Bengal Ir. C., Shumshabad. 18th Bengal Ir. C., Peshawur. 1st Gwalior Contingent Cavalry, Gwalior.
- 30. 2d Gwalior Contingent Cavalry, Augur. 1st Punjaub Cavalry, Dera Ismael. 2d Punjaub Cavalry, Dera Ismael. 3d Punjaub Cavalry, Bunnoo. 4th Punjaub Cavalry, Kohat. 5th Punjaub Cavalry, Asnee. 1st Oude Irregular Cavalry, Secrora. 2d Oude Irregular Cavalry, Lucknow. 3d Oude Irregular Cavalry, Pertabghur. Nagpoor Irregular Cavalry, Taklee. European Infantry. 8th Ft. (Qun.’s), Cawnpore. 10th Ft. (Qun.’s), Wuzeerabad. 24th Ft. (Qun.’s), Sealkote. 27th Ft. (Qun.’s), Sealkote. 29th Ft. (Qun.’s), Thayet Mhow. 32d Ft. (Qun.’s), Kussowlie. 35th Ft. (Qun.’s), Calcutta. 52d Ft. (Qun.’s), Lucknow. 53d Ft. (Qun.’s), Dugshai. 60th Ft. (Qun.’s), Jullundur. 61st Ft. (Qun.’s), Wuzeerabad. 70th Ft. (Qun.’s), Ferozpore. 75th Ft. (Qun.’s), Rawul Pindee. 81st Ft. (Qun.’s), Lahore. 87th Ft. (Qun.’s), Peshawur. 1st Europeans (East India Company’s), Dugshai. 2d Europeans (East India Company’s), Umballa. 3d Europeans (East India Company’s), Agra.
- 31. Native Regular Infantry. 1st Regiment, Cawnpore. 2d[39] Regiment, Barrackpore. 3d Regiment, Phillour. 4th Regiment, Noorpore. 5th Regiment, Umballa. 6th Regiment, Allahabad. 7th Regiment, Dinapoor. 8th Regiment, Dinapoor. 9th Regiment, Allygurh. 10th Regiment, Futteghur. 11th Regiment, Allahabad. 12th Regiment, Nowgong and Jhansi. 13th Regiment, Lucknow. 14th Regiment, Moultan. 15th Regiment, Meerut. 16th[39] Regiment, Meean Meer. 17th Regiment, Goruckpore. 18th Regiment, Bareilly. 19th Regiment, Berhampore. 20th Regiment, Meerut. 21st Regiment, Peshawur. 22d Regiment, Fyzabad. 23d Regiment, Mhow. 24th Regiment, Peshawur. 25th Regiment, Thayet Mhow. 26th Regiment, Meean Meer. 27th Regiment, Peshawur. 28th Regiment, Shahjehanpoor.
- 32. 29th Regiment, Jullundur. 30th Regiment, Agra. 31st Regiment, Barrackpore. 32d Regiment, Sonthal. 33d Regiment, Hosheapore. 34th Regiment, Barrackpore. 35th Regiment, Sealkote. 36th[40] Regiment, Jullundur. 37th[40] Regiment, Benares. 38th[41] Regiment, Delhi. 39th[41] Regiment, Jelum. 40th[41] Regiment, Dinapoor. 41st Regiment, Seetapoor. 42d Regiment, Saugor. 43d Regiment, Barrackpore. 44th Regiment, Agra. 45th Regiment, Ferozpore. 46th Regiment, Sealkote. 47th[41] Regiment, Prome. 48th Regiment, Lucknow. 49th Regiment, Meean Meer. 50th Regiment, Nagode. 51st Regiment, Peshawur. 52d Regiment, Jubbulpoor. 53d Regiment, Cawnpore. 54th Regiment, Delhi. 55th Regiment, Nowsherah. 56th Regiment, Cawnpore. 57th Regiment, Ferozpore. 58th Regiment, Rawul Pindee.
- 33. 59th Regiment, Umritsir. 60th Regiment, Umballa. 61st Regiment, Jullundur. 62d Regiment, Moultan. 63d Regiment, Barrackpore. 64th Regiment, Peshawur. 65th[41] Regiment, Dinapoor. 66th[42] Regiment, Almora. 67th[41] Regiment, {Etawah. {Minpooree. 68th Regiment, Bareilly. 69th Regiment, Moultan. 70th Regiment, Barrackpore. 71st Regiment, Lucknow. 72d Regiment, Agra. 73d Regiment, Jumalpore. 74th Regiment, Cawnpore. Irregular and Local Infantry. 1st Oude Irregular Infantry, Persadpore. 2d Oude Irregular Infantry, Secrora. 3d Oude Irregular Infantry, Gonda. 4th Oude Irregular Infantry, Lucknow. 5th Oude Irregular Infantry, Durriabad. 6th Oude Irregular Infantry, Fyzabad. 7th Oude Irregular Infantry, Lucknow. 8th Oude Irregular Infantry, Sultanpore. 9th Oude Irregular Infantry, Seetapoor. 10th Oude Irregular Infantry, Mullaong. 1st Gwalior Contingent Infantry, Gwalior. 2d Gwalior Contingent Infantry, Gwalior.
- 34. 3d Gwalior Contingent Infantry, Gwalior. 4th Gwalior Contingent Infantry, Gwalior. 5th Gwalior Contingent Infantry, Seepree. 6th Gwalior Contingent Infantry, Lullutpore. 7th Gwalior Contingent Infantry, Augur. 1st Punjaub Infantry, Kohat. 2d Punjaub Infantry, Kohat. 3d Punjaub Infantry, Kohat. 4th Punjaub Infantry, Dera Ghazi. 5th Punjaub Infantry, Bunnoo. 6th Punjaub Infantry, Dera Ismael. 1st Sikh Infantry, Hazara. 2d Sikh Infantry, Kangra. 3d Sikh Infantry, Khan. 4th Sikh Infantry, Umballa. 1st Nagpoor Irregular Infantry, Seetabuldee. 2d Nagpoor Irregular Infantry, Chandah. 3d Nagpoor Irregular Infantry, Raypoor. Regiment of Guides (foot and horse), Peshawur. Regiment of Kelat-i-Ghilzi, Shubkuddur. Regiment of Loodianah (Sikhs), Benares. Regiment of Ferozpore (Sikhs), Mirzapore. Ramgurh Light Infantry, Dorunda. Hill Rangers, Bhagulpore. Nusserree Rifles, Simla. Pegu Light Infantry, Myan Owng. Sirmoor Rifles, Almora. Kumaon Battalion, Deyra. Assam Light Infantry, 1st, Debroogurh. Assam Light Infantry, 2nd Gowhatti.
- 35. Mhairwarra Battalion, Bewar. Aracan Battalion, Akyab. Hurrianah Light Infantry, Hansi. Silhet Light Infantry, Cherrah. Malwah Bheel Corps, Sirdarpore. Mewar Bheel Corps, Khairwarah. Sebundee Corps, Darjeeling. Artillery, Engineers, Sappers and Miners. Horse-artillery, 1st Brigade: 3 European Troops. } 2 Native Troops. } Head-quarters: Horse-artillery, 2d Brigade: } Meerut. 3 European Troops. } Jullundur. 1 Native Troop. } Peshawur. Horse-artillery, 3d Brigade: } Umballa. 3 European Troops. } Cawnpore. 1 Native Troop. } Sealkote. Foot-artillery, 6 European Battalions. } Dumdum. (4 Companies each.) } Foot-artillery, 3 Native Battalions. } (6 Companies each.) } Engineers, } Head-quarters: Sappers and Miners, 8 Companies, } Roorkee. Mixed Corps—Cavalry, Infantry, and Artillery. Shekhawuttie Battalion, Midnapore. Jhodpore Legion, Erinpoora.
- 36. Malwah Contingent, Mehidpore. Bhopal Contingent, Sehore. Kotah Contingent, Kurrowlee. REGIMENTS AND STATIONS OF MADRAS ARMY—MAY 1857. Sir Patrick Grant, Commander-in-chief. European Cavalry. 12th Lancers (Queen’s), Madras. Native Cavalry. 1st Madras Light Cavalry, Trichinopoly. 2d Madras Light Cavalry, Sholapore. 3d Madras Light Cavalry, Bangalore. 4th Madras Light Cavalry, Kamptee. 5th Madras Light Cavalry, Bellary. 6th Madras Light Cavalry, Jaulnah. 7th Madras Light Cavalry, Secunderabad. 8th Madras Light Cavalry, Bangalore. European Infantry. 74th Foot (Queen’s), Madras. 84th Foot (Queen’s), Burmah.[43] 1st Europeans (East India Company’s), [Persia]. 2d Europeans (East India Company’s), Burmah. 3d Europeans (East India Company’s), Secunderabad. Native Infantry. 1st Regiment,[44] Secunderabad.
- 37. 2d Regiment, Quilon. 3d Regiment, Cananore. 4th Regiment, Burmah. 5th[44] Regiment, Berhampore. 6th Regiment, Burmah. 7th Regiment, Moulmein. 8th Regiment, Rangoon. 9th Regiment, Samulcottah. 10th Regiment, Rangoon. 11th Regiment, Cananore. 12th Regiment, Madras. 13th Regiment, Moulmein. 14th Regiment, Singapore. 15th Regiment, Burmah. 16th[44] Regiment, Mangalore. 17th Regiment, Madras. 18th Regiment, Madras. 19th Regiment, Bangalore. 20th Regiment, French Rocks. 21st Regiment, Paulghaut. 22d Regiment, Secunderabad. 23d Regiment, Russelcondah. 24th[44] Regiment, Secunderabad. 25th Regiment, Trichinopoly. 26th[44] Regiment, Kamptee. 27th Regiment, Vellore. 28th Regiment, Hosungabad. 29th Regiment, Penang. 30th Regiment, Cuddapah. 31st Regiment, Vizianagram.
- 38. 32d Regiment, Kamptee. 33d Regiment, Kamptee. 34th Regiment, Trichinopoly. 35th Regiment, Hurryhur. 36th[44] Regiment, Madras. 37th[45] Regiment, Burmah. 38th[44] Regiment, Singapore. 39th Regiment, Madras. 40th Regiment, Cuttack. 41st Regiment, Secunderabad. 42d Regiment, Secunderabad. 43d Regiment, Vizagapatam. 44th Regiment, Burmah. 45th Regiment, Rangoon. 46th Regiment, Henzana. 47th Regiment, Bellary. 48th Regiment, Moulmein. 49th[44] Regiment, Secunderabad. 50th Regiment, Bangalore. 51st Regiment, Pallamcottah. 52d Regiment, Mercara. Artillery, Engineers, Sappers and Miners. Horse-artillery, 4 European Troops. } Horse-artillery, 2 Native Troops. } Head-quarters: Foot-artillery, 4 European Battalions, (4 Companies each.) } } St Thomas’s Mount, Bangalore,
- 39. Foot-artillery, 1 Native Battalion. (6 Companies.) } } Kamptee, Saugor, Secunderabad. Engineers, Head-quarters: Fort St George. Sappers and Miners, Head-quarters: Dowlaishweram. REGIMENTS AND STATIONS OF BOMBAY ARMY—MAY 1857. Sir Henry Somerset, Commander-in-chief. European Cavalry. 14th Light Dragoons (Queen’s), Kirkee. Native Regular Cavalry. 1st Lancers, Nuseerabad. 2d Light Cavalry, Rajcote. 3d Light Cavalry, [Persia.] Native Irregular Cavalry. 1st Sinde Irregular Horse, Jacobabad. 2d Sinde Irregular Horse, Jacobabad. Poonah Irregular Horse, [Persia.] Gujerat Irregular Horse, Ahmedabad. South Mahratta Irregular Horse, [Persia.] Cutch Irregular Horse, Bhooj. European Infantry. 64th Foot (Queen’s), [Persia.] 78th Foot (Queen’s), Poonah.
- 40. 86th Foot (Queen’s), Kurachee. 1st Fusiliers (East India Company’s), Kurachee. 2d Light Infantry (East India Company’s), [Persia.] 3d Light Infantry (East India Company’s), Poonah. Native Regular Infantry. 1st Regiment,[46] Baroda. 2d[46] Regiment, Ahmedabad. 3d Regiment, Sholapore. 4th[47] Regiment, [Persia.] 5th Regiment, Bombay. 6th Regiment, Poonah. 7th Regiment, Poonah. 8th Regiment, Baroda. 9th Regiment, Surat. 10th Regiment, Nuseerabad. 11th Regiment, Bombay. 12th Regiment, Deesa. 13th Regiment, Hydrabad. 14th Regiment, Kurachee. 15th Regiment, Bombay. 16th Regiment, Shikarpore. 17th Regiment, Bhooj. 18th Regiment, [Aden.] 19th Regiment, Mulligaum. 20th Regiment, [Persia] 21st Regiment, Neemuch. 22d Regiment, Satara. 23d Regiment, [Persia.] 24th Regiment, Ahmednuggur.
- 41. 25th Regiment, Ahmedabad. 26th Regiment, [Persia.] 27th Regiment, Kolapore. 28th Regiment, Dharwar. 29th Regiment, Belgaum. Native Irregular Infantry. 1st Belooch Battalion, Kurachee. 2d Belooch Battalion, [Persia.] Khandeish Bheel Corps, Dhurrungaum. Rutnagherry Rangers, Rutnagherry. Sawunt Waree Corps, Sawunt Waree. Satara Local Corps, Satara. Kolapore Infantry Corps, Kolapore. Artillery, Engineers, Sappers and Miners. Horse- artillery, 1 European Brigade. } (4 Troops.)[48] } Head-quarters: Foot-artillery, 2 European Battalions. } Bombay. (4 Companies each.) } Ahmedabad. Foot-artillery, 2 Native Battalions. } Ahmednuggur. (6 Companies each.) } Engineers, Head-quarters: Bombay, Sappers and Miners, Head-quarters: Poonah and Aden.
- 42. Jumma Musjid, Agra.—Mosque built by Shah Jehan in 1656. 34. Presidency. Queen’s Regiments. Company’s Regiments. Total. Bengal, 16 3 19 Madras, 4 3 7 Bombay, 4 3 7 24 9 33 35. Presidency. Queen’s Regiments. Company’s Regiments. Total. Bengal, 15 4 19 Madras, 5 4 9 Bombay, 4 3 7 24 11 35
- 43. 36. First Division, under Major-general Stalker— Natives, 3550 Europeans, 2270 ———— 5820 Second Division, under Brigadier-general Havelock— Natives, 4370 Europeans, 1770 ———— 6140 37. In August 1857, of the whole railway distance marked out from Alexandria through Cairo to Suez, 205 miles in length, about 175 miles were finished—namely, from Alexandria to the crossing of the Nile, 65 miles; from the crossing of the Nile to Cairo, 65 miles; from Cairo towards Suez, 45 miles. The remainder of the journey consisted of 30 miles of sandy desert, not at that time provided with a railway, but traversed by omnibuses or vans. 38. ‘According to existing regulations of some years’ standing, every soldier on his arrival in India is provided with the following articles of clothing, in addition to those which compose his kit in this country: ‘Mounted Men.—4 white jackets, 6 pair of white overalls, 2 pair of Settringee overalls, 6 shirts, 4 pair of cotton socks, 1 pair of white braces. ‘Foot-soldiers.—4 white jackets, 1 pair of English summer trousers, 5 pair of white trousers, 5 white shirts, 2 check shirts, 1 pair of white braces. ‘These articles are not supplied in this country, but form a part of the soldier’s necessaries on his arrival in India, and are
- 44. composed of materials made on the spot, and best suited to the climate. ‘During his stay in India, China, Ceylon, and at other hot stations, he is provided with a tunic and shell-jacket in alternate years; and in the year in which the tunic is not issued, the difference in the value of the two articles is paid to the soldier, to be expended (by the officer commanding) for his benefit in any articles suited to the climate of the station. ‘The force recently sent out to China and India has been provided with white cotton helmet and forage-cap covers. ‘Any quantity of light clothing for troops can be procured on the spot in India at the shortest notice.’ 39. Grenadiers. 40. Volunteers. 41. Volunteers. 42. Goorkhas. 43. Removed to Calcutta. 44. Rifles. 45. Grenadiers.
- 45. CHAPTER XIV. THE SIEGE OF DELHI: JUNE AND JULY. hile these varied scenes were being presented; while sepoy regiments were revolting throughout the whole breadth of Northern India, and a handful of British troops was painfully toiling to control them; while Henry Lawrence was struggling, and struggling even to death, to maintain his position in Oude; while John Lawrence was sagaciously managing the half-wild Punjaub at a troublous time; while Wheeler at Cawnpore, and Colvin at Agra, were beset in the very thick of the mutineers; while Neill and Havelock were advancing up the Jumna; while Canning was doing his best at Calcutta, Harris and Elphinstone at Madras and Bombay, and the imperial government at home, to meet the trying difficulties with a determined front—while all this was doing, Delhi was the scene of a continuous series of operations. Every eye was turned towards that place. The British felt that there was no security for their power in India till Delhi was retaken; the insurgents knew that they had a rallying-point for all their disaffected countrymen, so long as the Mogul city was theirs; and hence bands of armed men were attracted thither by antagonistic motives. Although the real siege did not commence till many weary weeks had passed, the plan and preparations for it must be dated from the very day when the startling news spread over India that Delhi had been seized by rebellious sepoys, under the auspices of the decrepit, dethroned, debauched representative of the Moguls.
- 46. It was, as we have already seen (p. 70), on the morning of Monday the 11th of May, that the 11th and 20th regiments Bengal native infantry, and the 3d Bengal cavalry, arrived at Delhi after a night- march from Meerut, where they had mutinied on the preceding evening. At Delhi, we have also seen, those mutineers were joined by the 38th, 54th, and 74th native infantry. It was on that same 11th of May that evening saw the six mutinous regiments masters of the imperial city; and the English officers and residents, their wives and children, wanderers through jungles and over streams and rivers. What occurred within Delhi on the subsequent days is imperfectly known; the few Europeans who could not or did not escape were in hiding; and scanty notices only have ever come to light from those or other sources. A Lahore newspaper, three or four months afterwards, gave a narrative prepared by a native, who was within Delhi from the 21st of May to the 23d of June. Arriving ten days after the mutiny, he found the six regiments occupying the Selimgurh and Mohtabagh, but free to roam over the city; where the sepoys and sowars, aided by the rabble of the place, plundered the better houses and shops, stole horses from those who possessed them, ‘looted’ the passengers who crossed the Jumna by the bridge of boats, and fought with each other for the property which the fleeing British families had left behind them. After a few days, something like order was restored, by leaders who assumed command in the name of the King of Delhi. This was all the more necessary when new arrivals of insurgent troops took place, from Allygurh, Minpooree, Agra, Muttra, Hansi, Hissar, Umballa, Jullundur, Nuseerabad, and other places. The mutineers did not, at any time, afford proof that they were really well commanded; but still there was command, and the defence of the city was arranged on a definite plan. As at Sebastopol, so at Delhi; the longer the besiegers delayed their operations, the greater became the number of defenders within the place, and the stronger the defence-works. It must be remembered, in tracing the history of the siege of Delhi, that every soldier necessary for forming the siege-army had to be brought from distant spots. The cantonment outside the city was
- 47. wholly in the hands of the rebels; and not a British soldier remained in arms in or near the place. Mr Colvin at Agra speedily heard the news, but he had no troops to send for the recapture. General Hewett had a British force at Meerut—unskilfully handled, as many persons thought and still think; and it remained to be seen what arrangements the commander-in-chief could make to render this and other forces available for the reconquest of the important city. Major-general Sir Henry Barnard was the medium of communication on this occasion. Being stationed at Umballa, in command of the Sirhind military division, he received telegraphic messages on the 11th of May from Meerut and Delhi, announcing the disasters at those places. He immediately despatched his aid-de-camp to Simla, to point out the urgent need for General Anson’s presence on the plains instead of among the hills. Anson, hearing this news on the 12th, first thought about his troops, and then about his own movements. Knowing well the extreme paucity of European regiments in the Delhi and Agra districts, and in all the region thence eastward to Calcutta, he saw that any available force to recover possession of Delhi must come chiefly from Sirhind and the Punjaub. Many regiments were at the time at the hill-stations of Simla, Dugshai, Kussowlie, Deyrah Dhoon, Subathoo, c., where they were posted during a time of peace in a healthy temperate region; but now they had to descend from their sanitaria to take part in stern operations in the plains. The commander-in-chief sent instant orders to transfer the Queen’s 75th foot from Kussowlie to Umballa, the 1st and 2d Bengal Europeans from Dugshai to Umballa, the Sirmoor battalion from Deyrah Dhoon to Meerut, two companies of the Queen’s 8th foot from Jullundur to Phillour, and two companies of the Queen’s 81st foot, together with one company of European artillery, from Lahore to Umritsir. These orders given, General Anson himself left Simla on the evening of the 14th, and arrived at Umballa early on the 15th. Before he started, he issued the proclamation already adverted to, announcing to the troops of the native army generally that no cartridges would be brought into use against the conscientious wishes of the soldiery; and after he arrived at Umballa,
- 48. fearing that his proclamation had not been strong enough, he issued another, to the effect that no new cartridges whatever should be served out—thereby, as he hoped, putting an end to all fear concerning objectionable lubricating substances being used; for he was not aware how largely hypocrisy was mixed up with sincerity in the native scruples on this point. Anson and Barnard, when together at Umballa, had to measure well the forces available to them. The Umballa magazines were nearly empty of stores and ammunition; the artillery wagons were in the depôt at Phillour; the medical officers dreaded the heat for troops to move in such a season; and the commissariat was ill supplied with vehicles and beasts of burden and draught. The only effectual course was found to be, that of bringing small detachments from many different stations; and this system was in active progress during the week following Anson’s arrival at Umballa. On the 16th, troops came into that place from Phillour and Subathoo. On the 17th arrived three European regiments from the Hills,[49] which were shortly to be strengthened by artillery from Phillour. The prospect was not altogether a cheering one, for two of the regiments at the station were Bengal native troops (the 5th and 60th), on whose fidelity only slight reliance could be placed at such a critical period. In order that no time might he lost in forming the nucleus of a force for Delhi, some of the troops were despatched that same night; comprising one wing of a European regiment, a few horse, and two guns. On successive days, other troops took their departure as rapidly as the necessary arrangements could be made; but Anson was greatly embarrassed by the distance between Umballa and the station where the siege-guns were parked; he knew that a besieging army would be of no use without those essential adjuncts; and it was on that account that he was unable to respond to Viscount Canning’s urgent request that he would push on rapidly towards Delhi. On the 23d of May, Anson sketched a plan of operations, which he communicated to the brigadiers whose services were more immediately at his disposal. Leaving Sir Henry Barnard in command
- 49. Welcome to our website – the ideal destination for book lovers and knowledge seekers. With a mission to inspire endlessly, we offer a vast collection of books, ranging from classic literary works to specialized publications, self-development books, and children's literature. Each book is a new journey of discovery, expanding knowledge and enriching the soul of the reade Our website is not just a platform for buying books, but a bridge connecting readers to the timeless values of culture and wisdom. With an elegant, user-friendly interface and an intelligent search system, we are committed to providing a quick and convenient shopping experience. Additionally, our special promotions and home delivery services ensure that you save time and fully enjoy the joy of reading. Let us accompany you on the journey of exploring knowledge and personal growth! ebookultra.com