Medicine as a data science

1. Intereg Project
Biomedical Informatics
Ljiljana Majnarić Trtica
I. Medicine as a data science

2. Medicine - a data science
 In the last decades, the life science, biomedicine and health care - are increasingly
turning into a data intensive science
 This is associated with the expansion of available electronic data, including:
 Digitization of electronic health records (EHR)
 Aggregation of research data into databases in pharmaceutical industries
 The release of stored patient data by the governments for research purposes (e.g.
patients health insurance claims)
 Aggregation of research data from clinical trials, epidemiological and biomedical
research
 The emergency of the high tech medicine (omics-medicine)
 The onset of patient self-tracking and remote monitoring by using mobile devices
and biosensors

3. The classical definition of Medical
Informatics
Decision Support is the key
role of Medical Informatics

4. Fig. 1
Biomedical Informatics vs Medical Informatics

5. Characteristics of data in biomedicine
 Different resources in biomedicine continue to actively generate data of
 increasing size (volume) and diversity (from different resources)
 Characteristics of this data are also that they are:
 Multi-dimensional (of different meaning, subclasses)
 Highly complex (e.g. the microscopic structure of a yeast protein network) (Fig. 2)
 Often weakly structured (as the text in the patient records, the signals from physiological
sensors)
 Noisy (missing and inconsistent)
 There is a growing need for integrative analysis and modeling of this data

6. Fig. 2.
The computer-based visualization of the yeast protein network
 A great challenge is to find unknown structures (structural homologies) among the
enormous set of uncharacterized data
 By application of a special visualization method, such structures can be made
graphically visible, enabling medical professionals to understand these data more
easily

7. Big Data
 Increase in volume and diversity of data in biomedical practice and science - is
collectively termed ” Big Data”
 Big Data represent a unique opportunity to gain insights, derive knowledge and
foster discovery that will result in improved patient outcomes, reduced costs and
accelerated biomedical advances

8. Examples of Big Data usability in medical practice
 To boost the applicability of clinical research studies into real-world scenarios
where population heterogeneity is an obstacle
 (thus, changing the paradigm - from the hypothesis driven to data-driven
medicine) (Fig. 3)
 To foster extraction and effective and innovative use of knowledge hidden within
the huge volumes of data
 To enable patients identification who are at risk for unfavorable health outcomes
(disease, death, hospital (re-)admission)
 To enable effective and precision medicine through patient stratification - a key
task toward personalized health care
 To enable predicitive analytics in personalized health care (Fig. 5)

9. 1) Hypothesis-driven vs data-driven medicine
2) Clinical research tasks should determine research methods - opposite to what is
nowadays -
where clinical projects meet the criteria of the established research methods
3) From descriptive (curable) to predictive (preventable) and proactive, participatory (with
patient active participation) medicine
4) From customized (one-size-fits-all) to personilized prescriptions
Big Data and Paradigm changes in biomedical science

10. Big Data and Omics-based medicine
 The increasing interest for personalized medicine evolves together with two
major technological advances
 First, the new-generation, rapid and less expensive, DNA sequencing method
 combined with the remarkable progresses in molecular biology - leading to the
post-genomic era (transcriptomics, proteomics, metabolomics)
 Second, the refinement of computing tools, which allows the immediate analysis of
a huge amount of data
 thus, creates a new universe for medical research, that of „big data”, when
analyzed by computerized modelling

11. Fig. 3
The challlenge of Big Data analysis in omics-based medicine
Genomics, proteomics, metabolomics
Genes and molecular pathways and networks
Systems biology - integrative analysis of data of different levels of bodily
organization

12. (P4) medicine - Personalized,
Predictive, Preventive and Participatory medicine
Key benefits of P4 medicine
The ability to
- detect disease at an earlier stage, when it is easier
and less expensive to treat effectively
- stratify patients into groups that enable the selection
of optimal therapy (Fig. 5)
- reduce adverse drug reactions by more effective
early assessment of individual drug responses
- improve the selection of new biochemical targets
for drug discovery
- reduce the time, cost and failure rate of clinical trials
for new therapies
- shift the emphasis in medicine from reaction
to prevention
and from disease to wellness
Fig. 5. Stratification of patients into groups
to enable the selection of optimal therapy

13. Obstacles
to effectively use Big Data for practical purposes
 The problem of heterogeneous data (biomedical data result from various sources and different
structural dimensions - varying from microscopic world (e.g. omics-data)to the macroscopic world
(e.g. disease spread in the population of the public health informatics)
 The problem of data sharing and distribution among different providers and departments
 Often noisy, missing, inconsistent and non-standardized data
 There is a gap between the ”natural”, available data, and the data applicable for practical purposes
the machinery and procedures for data processing

14. Knowledge discovery in databases (KDD)
 A classical interpretation of the KDD process includes several steps: data selection, pre-
processing, transforming, data mining and interpretation (Fig. 6)
Fig. 6. The overview of steps that encompass the KDD process

15. Knowledge discovery from Big Data
 The challenge is to
 extract meaningful information from data
 gain new knowledge
 discover previously unknown insights
 look for patterns
 make sense of data

16. Knowledge discovery from Big Data
 Many different approaches
 New mathematical and graphical methods
 Data Mining (DM)and Machine Learning (ML) methods - mostly used methods in
the past
 Data mining - a key step in the KDD process the term Knowledge Discovery
and Data Mining (KDD)

17. Data Mining
 The computational process of discovering previously unknown, valid patterns and
relationships in large data sets - for prediction, classification and clustering
purposes
 involving combinations of sophisticated methods
 Statistical models
 Mathematical algorithms
 ML methods - algorithms that improve their performance automatically through
experience
 Has led to the development of

18. Application of Data Mining Techniques
in Health Care Domain
 Has led to
 the developmnet of intelligent systems and decision support systems (rule-based expert
systems)
 improvement of the prediction of unfavorable health outcomes and diagnosis
 improved disease classification
 the discovery of relationships between pathological data and clinical data and between
patients characteristics and medications efficiency
 candidate selection process for medical tests and procedures

19. The concept of the Human-Computer Interaction (HCI)
 The Knowledge Discovery - a process ranging from the physical side of data to the
human side of knowledge (defined as the cognitive process)
 The challenge is in making knowledge to be usable by the end users (by making
sense of data)
 The process added to KDD is INTERACTION (COMMUNICATION) with the
human end user (medical expert)
 It is the human end user (not machine) who posses the problem solving
intelligence, hence, the ability to ask intelligent questions about the data
 The human (medical expert) is able to solve complex problems sometimes
intuitively (that is, without the need to describe the exact rules or processes used
during the problem analysis)

20. Interaction is the key topic
in the concept of HCI
A novel approach is to combine KDD & HCI

21. ...or as Albert Einstein said
US (German-born) physicist (1879 - 1955)
Computers are incredibly fast, accurate, and stupid.
Human beings are incredibly slow, inaccurate, and
brilliant.
Together they are powerful beyond imagination.