BUILDING TODAY THE FUTURE CADRE

OF INFORMATICS-ORIENTED PHYSICIANS

Lessons from a New Biomedical Informatics Curriculum for Medical Doctors

Ronen Tal-Botzer

1,*

, Rachel S. Levy-Drummer

1,*

, Gidi Rechavi

and Ron Unger

The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel

Center for Cancer Research, Sheba Medical Center, Tel Hashomer, Ramat-Gan, Israel

These authors contributed equally to this work

Keywords: Biomedical informatics, Bioinformatics, Data mining, Medical education, Curriculum.

Abstract: Recent revolutions in bioinformatics, web and smartphone apps, and the initiative to significantly increase

the implementation of electronic medical records, highlighted the need of physicians to be familiar with the

newly emerging field of Biomedical Informatics.The huge potential of the field motivated us to educate a

cadre of physicians with the required informatics skills, and consequently, improve the quality of medical

care in Israel. Here we describe the curriculum of our certification study program in Biomedical Informatics

for physicians, comprised of the following courses: Advanced Molecular Methods for Clinical Applications,

Clinical Bioinformatics, Biomathematical Modelling, Biostatistics and the Design of Clinical Trials,

Clinical Systems Biology and Medical Data Mining. We discuss the rationale of the program; explain our

considerations in designing the curriculum; describe the content of each course; highlight the Medical Data

Mining course which was specifically designed for this program; and discuss the feedback from the

students. More information is available at the website of the program: www.bio-medical.info.

1 FORMATION

OF A NEW DISCIPLINE

The interface between medicine and computer

science has long been inspiring both researchers and

clinicians from the two disciplines. Nevertheless,

three recent revolutions, just from about the past five

years, have led us to realize that the time has come

to initiate a curriculum in Biomedical Informatics

dedicated to students, who already have a Medical

Doctor degree, and have worked as physicians for

several years already. These revolutions, as we see

it, are:

 Bioinformatics: The recent flourish in

Bioinformatics databases and research, thanks

to next generation sequencing technologies.

 Web & Mobile: The amazing popularity and

intensive usage of social networks and

smartphones in almost every aspect of our

lives, including medical treatment and

research.

 EMRs: The US health care stimulus package

and funds of 2009 for physicians to use

Electronic Medical Records (EMRs) and the

construction of the National Health

Information Network (NHIN), that will bring

enormous medical datasets together for the

first time.

The new study program, which began in 2010

and is now during its second year of operation, is a

joint venture of Bar-Ilan University and the Sheba

Medical Center in Israel. Its goal is to elucidate

medical doctors' perspectives regarding the

biomedical informatics revolution, as well as to

provide them with both the theoretical basis and

practical skills to enhance their medical practice and

research with computational tools. Since the

program is new and we consider it as a “pilot”,

during its first two years of operation we decided to

accept around one dozen students per year.

Bar-Ilan University was the first university in

Israel, and probably one of the first ones in the

world, to establish undergraduate and graduate study

programs in Computational Biology already 15

years ago. Almost all universities in Israel have

since opened similar programs, contributing to

Israel’s success in Bioinformatics.

371

Tal-Botzer R., Levy-Drummer R., Rechavi G. and Unger R..

BUILDING TODAY THE FUTURE CADRE OF INFORMATICS-ORIENTED PHYSICIANS - Lessons from a New Biomedical Informatics Curriculum for

Medical Doctors.

DOI: 10.5220/0003887603710377

In Proceedings of the International Conference on Bioinformatics Models, Methods and Algorithms (SSTB-2012), pages 371-377

ISBN: 978-989-8425-90-4

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

The Sheba Medical Center is the largest hospital

and medical research facility in Israel. Both Sheba

Medical Center and Bar-Ilan University are located

in the city of Ramat-Gan, a fact that enables very

fruitful cooperation between researchers and trainers

from the two institutes, as well as convenience for

Sheba's physicians to join the study program at Bar-

Ilan's campus.

The program was approved by the Israeli

Medical Association (IMA), which accredits it as

“continuing education” for physicians. The complete

information about the program, as well as details for

prospective students, are available at its website:

www.bio-medical.info.

2 THE NEED FOR

INFORMATICS-ORIENTED

PHYSICIANS

When building the new study program we first tried

to understand what kind of new challenges future

physicians and researchers will be required to face.

The three above-mentioned revolutions

(Bioinformatics, Web & Mobile and EMRs) are

already changing the very basic foundations of

medical treatment and of medical research (Stead et

al., 2010). Following are just few examples.

2.1 The Change in Medical Treatment

A patient today is much more informed about his or

her health condition and treatment options than he or

she used to be just a decade ago. More and more

patients consume medical information about their

symptoms, diagnostic methods, diseases,

medications, side effects, alternative treatment

options, potential physicians, etc., from either

reliable or unreliable sources online.

Not only are health related search queries and

websites gaining more popularity than ever before

(Schmidt, 2008), but also unstructured and

uncontrolled information is exchanged through

social networks (Dreesman and Denecke, 2011; and

Roblin, 2011). The “wisdom of the crowd” is now

being formalized in web services like “Patients Like

Me”, where patients voluntarily report their health

condition on a daily basis. This information is later

being analyzed and sliced per disease per symptom

per treatment per side effect (Wicks, 2011). Whether

physicians like or dislike this phenomenon (indeed,

many of them report their frustration from that

situation) this revolution is here to stay, and

physicians should be capable to react accordingly.

In addition to just better management of the

information, smartphone apps and auxiliary gadgets

help patients monitor health related signals on a 24/7

basis (Shetty, 2011), thus collecting more data and

different kinds of data than before. For example,

many people are using apps like the “Sleep Cycle”

alarm clock, that analyzes their movements during

sleep, in order to be woken up at optimal times,

when they are not in a deep sleep mode (REM),

hence waking up is easier.

But indisputably, the more interesting data is and

will be coming very soon from the Bioinformatics

and EMRs revolutions. Next generation sequencing

technologies and patients' medical histories will

drive medical treatment to the personalization era.

Computer science plays a crucial role in combining

the various information sources and kinds regarding

an individual patient, analyzing them and outputting

personalized diagnostics and recommended

therapeutic pathways. Obviously, when combining

the “official” genomic and medical history data with

“softer” information, like patient’s behavioural data,

collected from one’s mobile device or social

network, a much deeper medical picture of the

patient can be assessed (Shah and Robinson, 2011).

Interestingly enough, it is web algorithms, such

as those that personalize feeds in social networks or

deliver personalized ads to mobile devices, which

are considered most promising in enabling better

medical personalization and recommendations. A

good example would be the Netflix movie rental

web service, which is famous for its superior

algorithmic ability to predict personal preference in

movie selection.

2.2 The Change in Medical Research

Hand in hand with the above-mentioned advances in

personalized medicine goes the need for adequate

research and skilled researchers. The three

revolutions described in the first chapter

(Bioinformatics, Web & Mobile and EMRs) bring so

many new kinds of information, at higher

resolutions and much higher volumes, that they

actually constitute a brand new platform for medical

research by their own (Stead et al., 2010). The

information is so different from what researchers

used to know and work with, that it certainly holds a

huge potential for new kind of medical discoveries

and new innovative treatment solutions.

BIOINFORMATICS 2012 - International Conference on Bioinformatics Models, Methods and Algorithms

372

3 OTHER BIOMEDICAL

INFORMATICS EDUCATION

PROGRAMS

Clearly, due to this kind of revolution, it’s being

realized by many educators that there is a need to

offer suitable training programs for physicians

(Shortliffe, 2010). However, only very few medical

schools offer significant computer science and

informatics education within the course of regular

medical studies. This is understandable due to the

justified conservative nature of the medical

discipline and the heavy load of medical programs,

but also since these revolutions are quite recent. The

first personal genome (of James Watson) was

sequenced in 2007, The first iPhone being released

in 2007, Facebook being opened to the public in

2006 and the US stimulus package for EMRs being

legislated only in 2009.

Major medical schools and research institutes, like

Stanford University School of Medicine and

Columbia University Medical Center, have recently

added some mandatory and elective courses in

informatics related subjects. This is done in parallel to

dedicated programs, offered by the Stanford Center

for Biomedical Informatics Research

(bmir.stanford.edu) and by Colombia University’s

Department of Biomedical Informatics

(www.dbmi.columbia.edu/education). Both institutes

offer large-scale dedicated programs for

undergraduate and graduate students, which cover the

full theoretical background required in computer

science, biology and medicine. A thorough database

and reviews on such programs can be obtained

through the American Medical Informatics

Association (Academic Informatics Programs, 2011).

In addition, some institutes also offer

certification study programs for professionals,

similar to our program, with differences in program

focus, length and audience type. One of the most

important programs is given by the American

Medical Informatics Association (AMIA 10x10

Courses, 2011). Another notable curriculum of a

certification study program, similar to the one

described in this paper, is the University of Texas

Certificate Program of Health Informatics. However,

it is designed mainly for professionals working in

the healthcare and information technology (IT)

fields, rather than specifically to practicing

physicians.

Since Biomedical Informatics is a new emerging

field, it is not clear what should be included in the

curriculum. For an interesting discussion, see a

recent commentary in the Journal of American

Medical Association (JAMA), (Shortliffe, 2010).

Clearly, each institute has its own considerations

when setting up such programs. In many cases, these

are the result of specific fields of interest of current

faculty members and/or the attributes of the medical

services in each country. For example, Israel has

quite a unique structure of only four HMOs, which

combine the medical services together with the

medical insurances for the vast majority of the

population. The result is a relatively centralized

management system of medical information, as well

as standardized medical testing procedures and lab

equipment. This situation should facilitate the

emergence of the Biomedical Informatics field in

Israel.

4 COMPOSING THE

BIOMEDICAL INFORMATICS

CURRICULUM

4.1 Guidelines

Since current curricula in medical schools are

already too condensed and, as mentioned earlier,

conservative by nature, we realized there is no

realistic option to add significant computer science

and/or informatics study units to current medical

studies. A better option is to initiate a certification

study program for young medical doctors during

their resident years or as young attendants. This

segment of prospective students enjoys the

advantage of having real life experience as

physicians, has access to clinical databases and the

ability to apply their knowledge to the benefit of

their patients. These guidelines led us to build a

program that can fit into the schedule of practicing

physicians.

4.2 Program Scope

Based on a priori discussions with candidate

physicians, we understood that the way to achieve

optimal learning atmosphere with minimal

interference from other obligations physicians carry,

is to assemble the course meetings in one full day

per week. It also became clear that for practical

reasons, given the career path of young physicians,

the program duration cannot exceed the length of

one year. Thus, we ended up designing a program of

one day a week for two academic semesters. This

framework allowed us to offer six courses, three in

BUILDING TODAY THE FUTURE CADRE OF INFORMATICS-ORIENTED PHYSICIANS - Lessons from a New

Biomedical Informatics Curriculum for Medical Doctors

373

each semester. Each meeting of each course consists

of 1.5 hours of frontal lecture, followed by an hour

of hands-on and exercise session. Each semester

lasts 14 weeks, therefore the whole curriculum spans

over 210 hours in class. Both theoretical and

programming assignments are given throughout the

semester, and exams or final projects are given by

the end of each course.

It is clear that within such limited time we cannot

train physicians to become experts in biomedical

informatics. Rather, our aim is to widen their

horizon; make them aware of the new technologies

and the new research and treatment opportunities;

make them familiar with the main concepts and tools

in the field; and enable them to collaborate

effectively with experts. These goals led us to design

a program that gives a broad perspective on the

emerging field of biomedical informatics, rather than

building a narrow program around one or two

specific topics. In this way, each student will have

enough background to continue further, if he or she

chooses to, and deepen his or her knowledge in

specific topics, if the need and opportunity arise.

It was clear that the program should start from

the body of knowledge that is by now considered

classic in the field, as a foundation to the more

advanced topics. Thus, our current curriculum

includes the following six courses: “Advanced

Molecular Methods for Clinical Applications”,

“Clinical Bioinformatics”, “Biomathematical

Modelling”, ”Biostatistics and the Design of Clinical

Trials”, “Clinical Systems Biology” and “Medical

Data Mining".

We had a dilemma whether we should include

computer programming as one of our subjects.

Clearly, mastering biomedical informatics requires

the ability to program and to grasp the fundamental

concepts of computer science. However, computer

programming is a “heavy” discipline, which requires

far more teaching and exercise time than we can

afford. Since, as we mentioned above, our goal was

to give the students a broad introduction to the field,

rather than make them practicing experts, we

decided not to devote a course to programming. In

order just to give a sense of programming, we teach

the Perl script language, as part of the exercise

sessions of the Clinical Systems Biology course.

4.3 Students’ Feedback and the

Resulting Changes in the Program

By the end of each course we asked the students to

fill a detailed questionnaire about their satisfaction

levels and asked for constructive suggestions. In

addition, by the end of the two semesters, we

gathered all students and faculty together for a joint

review of the past year.

It was not surprising to find out that some of the

topics, which we initially considered as very

relevant to physicians, easy to grasp, easy to

implement and/or just interesting, turned out to be

the opposite, and of course, vice versa. We used this

feedback, although somewhat diverse, to fine tune

the exact selection of topics in the following year.

We realized that there were opposing views

among the students themselves regarding the

balance between the theoretical/scientific elements

and the hands-on/ exercise elements. While some

students preferred the theoretical approach and said

that they can later find web resources to practice by

themselves, other students said that their main gain

was the introduction of tools that they can

implement in their own research.

As a result, we decided, first, to keep the hands-

on topics that we considered as either essential or

non-trivial. Such topics include the different

bioinformatics and systems biology algorithms and

tools, the SQL language, the Weka 3 suite; and more

(see detailed syllabus in Chapter 5). Second, we

decided to add to the existing curriculum a set of

“satellite” sessions, in which students are not

obligated to participate in. In these sessions we

included topics from the original program that we

considered as not being at the essential core of the

biomedical informatics profession. Such topics

include more challenging usage of the Perl script

language, telemedicine, text mining methods and

more. This differentiation of topics and lessons was

also positively accepted by the students of the

second class, and most of them showed interest in

taking the satellite courses.

5 PROGRAM TOPICS BY

COURSES

Below is a short description and notable insights

regarding five of the courses in our curriculum,

followed by a more detailed description of a sixth

special course, “Medical Data Mining”. Courses

similar to the first courses have been given in

various versions for several years already within our

Computational Biology programs for undergraduate

and graduate students, and the challenge was to

modify them towards clinical applications such that

they will be suitable to an audience of medical

doctors. However, the “Medical Data Mining”

BIOINFORMATICS 2012 - International Conference on Bioinformatics Models, Methods and Algorithms

374

course was built from scratch, specifically designed

to address the most recent advancements in this

field.

5.1 Clinical Bioinformatics

This is a fundamental course in many biomedical

informatics programs. The course introduces the

students with the major bioinformatics databases for

DNA, RNA and proteins, as well as more

specialized databases of diseases, miRNA and

microarray experiments. The computational methods

described include sequence alignment, database

searches, phylogenetic analysis, genome viewers,

protein secondary and tertiary structure prediction,

etc.. The clinical aspects are emphasized by using

examples of genes and proteins involved in diseases,

as well as explaining, for example, the flow of

analysis that can lead us from a mutation in the

genome to its clinical manifestation.

5.2 Biomathematical Modeling

This field of theoretical biology, which strives to

model biological systems and their kinetic behaviour

using differential equations, is usually not covered

within classic bioinformatics programs. This field is

very applicable to describe clinical situations, like

the progression of a disease in the individual

patient’s body and the interplay between viruses and

the immune system. Since we have the local

expertise to teach such a course, we decided to

include it in the program. A gentle reminder is given

to the required mathematical knowledge; and we

select examples where the level of the required

mathematics is suitable.

5.3 Advanced Molecular Methods for

Clinical Applications

This course covers new genomic technologies. It is a

bit different from the other courses, in the sense that

its main subjects are experimental methods rather

than computational ones. However, we realized that

physicians need a refreshment of their knowledge

about the basic concepts of molecular biology; and

especially need introduction to novel genomic

technologies that emerged recently. Thus, the course

presents the technologies of micro-arrays, deep

sequencing and ncRNA analysis; and discusses their

clinical applications, especially for cancer research.

The computational aspects of the analysis of the

results are highlighted.

5.4 Biostatistics and the Design of

Clinical Trials

This course starts with a reminder on the basic tools

in statistical analysis and moves to more advanced

topics, like the proper statistical design of Clinical

Trials. A special attention is given to the statistics of

the analysis of large scale data. When data mining

algorithms scan a huge number of hypotheses to

explain the data, the issue of False Discovery Rate

becomes much more critical than in regular

statistical analyses.

5.5 Clinical Systems Biology

Systems biology deals with the organization and

control of large and complex biological systems, like

organs, diseases and genomes that consist of

thousands of genes, proteins and metabolites.

Systems biology aims to explain biological

processes by the intricate networks of interactions

between these elements. The course deals with both

the theoretical aspects of the field and the practical

tools that have been developed. A special emphasis

is given to disease pathways and the possible

implications of systems biology to clinical

intervention.

5.6 Medical Data Mining

The use of computational methods to mine medical

data, information and knowledge, as it is diverse and

scattered among EMRs, medical journals, social

networks and the web in general, has long been one

of the most desired goals of Artificial Intelligence

(A.I.). The potential applications in real-life

situations and the implications to the medical

industry, if and when algorithms function as good as

we intend, are enormous and can hardly be grasped

or imagined (Kim et al., 2011). Thanks to the

growing spread of A.I. algorithms throughout the

web, and its close and quick connection to large

revenues (such as A.I. for online advertising,

ranking search results, recommendation engines in

e-commerce sites, etc.), data mining algorithms have

recently evolved, and in a tremendous scale.

More and more academic institutes, as well as

start-up companies, channel these advancements to

the benefit of the medical field. We realized that

although this subject is relatively new, and although

most of the actual tools are commercial, this

phenomenon should not be ignored by future

physicians. On the contrary, they should understand

the inherent limits and risks of such A.I. tools, as

BUILDING TODAY THE FUTURE CADRE OF INFORMATICS-ORIENTED PHYSICIANS - Lessons from a New

Biomedical Informatics Curriculum for Medical Doctors

375

well as to be able to identify the circumstances,

where such tools could be helpful, and might deliver

intelligent and educated conclusions, even more than

expert human physicians (Korhonen et al., 2009).

Another important goal is to provide physicians the

basic skills required to implement, tune and even

participate in the development process of such tools.

The course is built as a dialog between horizontal

and vertical subjects, being the medical applications

of data mining algorithms, on one hand, and the

computational methodologies that are intertwined in

each of those applications, on the other hand. The

topics regarding medical applications are: the

automatic collection and management of medical

data, information and knowledge; the semantic

analysis of text, the individual-based vs. the

population-based analysis of EMRs; the 4-dimension

visualization of biological phenomena based on even

sparse pieces of partial knowledge; the prediction of

treatment outcome; the decentralization of medical

assessment and treatment using telemedicine aided

encounters; and the overall movement towards an

era of a much more personalized medicine, which is

based, among other types of clinical data, on the

patient’s own genome.

The other axis of topics - the computational

methodologies, is comprised of the fundamental

theories of computation and data mining, as well as

basic algorithms in this field. The ones we found as

most relevant are: data representation options,

information retrieval (IR) engines, pattern

recognition, motif discovery, clustering, classifiers,

machine learning, multi-dimensional search,

decision trees and basic graph theory. In addition to

the theoretic lectures, we hold hands-on exercise

sessions, in which we teach the SQL language and

the ‘Weka 3’ data mining suite.

Two additional special chapters are dedicated to

smartphone apps for physicians, such as:

‘ePocrates’, ‘AOTrauma’ and ‘Calculate’; and to

medicine-oriented social networks and web services,

such as: ‘Patients Like Me’, ‘23andMe’ and

‘Knome’. Since these two topics can be considered

as somewhat more “stand alone”, not tightly

connected to the other more classic data mining

topics, they are taught in the framework of special

seminar. Most of the seminar is built around real-life

examples. They are demonstrated and discussed by a

practicing physician, who uses such applications in

his day-to-day work. In addition to our students, we

invite to this special seminar other Israeli

researchers, physicians and managers from the

medical domain, who still do not practice

biomedical informatics. We do that in order to

harbour these easily explained yet impressive topics

as an introduction to the field and our study program

and encourage enrolment for the following year.

6 DISCUSSION

Computational Biology is a result of the marriage of

Computer Science and Biology. Now that

computational biology has come of age, the time is

ripe for the next step of integrating computational

biology with medicine. Clearly, there are many

considerations when starting such a program and

many compromises and adjustments have to be

made in order to fit the program to the needs of the

students. One of the problems we encountered is the

diverse background and interest of the students, thus,

it was not simple to come up with a program that

would appeal to all students.

The feedback we received from the participants of

the first year of the program was very encouraging.

All students mentioned that the program opened their

eyes to the new world of research and treatment

opportunities. However, we could see again the

diversity among our students. While some of the

students wanted the program to be more hands-on and

practical, others actually suggested concentrating

more on theoretical issues. While some said they

preferred to get a stronger mathematical and

computational basis (including programming), other

students complained that this material was already

excessive and too difficult for them.

For us it was amazing to see how quickly the

program paid off. One of the students, a children

neurologist, told us that being inspired by the

Systems Biology course, she changed the treatment

of one of her patients. It is a patient who suffered

from the rare Ruvalcabra syndrome, which is caused

by mutations in the PTEN gene. The patient was

treated for seizures and behavioural problems

without much success. Led by insights she gained

from the course, she tried off-label treatment with

Rapamune, a drug that is often prescribed for

Tuberous Sclerosis, which is another syndrome that

shares the same pathway with Ruvalcabra.

Surprisingly, the patient showed a significant

cognitive improvement. While additional research is

needed to follow up this and other patients, it is clear

that the program already inspired clinical innovation.

A lot of research and collaborative thinking were

put together in this initiative. Our hopes are that

other institutes in Israel and around the world will

follow with similar programs, and that together we

will influence the building of the future cadre of

BIOINFORMATICS 2012 - International Conference on Bioinformatics Models, Methods and Algorithms

376

physicians. According to our own scientific research

experience, the exhibited emerging phenomenon in

the world and the extremely positive feedback we

received from physicians who graduated our

program, we have no doubt that this future cadre

will be much more informatics-oriented.

ACKNOWLEDGEMENTS

We thank Eran Eyal, Nineth Amariglio, Erez

Levenon, Avidan Neumann, Yanay Ofran, Ramit

Mehr, Sol Efroni, Hiba Waldman Ben-Asher and

Niv Marom for their role in designing the courses

and teaching in the program.

REFERENCES

Stead, WW., Searle, JR., Fessler, HE., Smith, JW.,

Shortliffe EH., 2010. Biomedical informatics:

changing what physicians need to know and how they

learn. In Academic Medicine. 86(4), 429-434.

Schmidt, E., 2008. Keynote speech. In The Healthcare

Information and Management Systems Society

Conference.

www.youtube.com/watch?v=dTZKNcx9sBA.

Dreesman, J., Denecke, K., 2011. Challenges for signal

generation from medical social media data. In Studies

in health technology and informatics. 169, 639-643.

Roblin, DW., 2011. The potential of cellular technology to

mediate social networks for support of chronic disease

self-management. In Journal of Health

Communication. Volume 16.

Wicks, P., Vaughan, TE., Massagli, MP., Heywood J.,

2011. Accelerated clinical discovery using self-

reported patient data collected online and a patient-

matching algorithm. In Nature Biotechnology. 29,

411–414.

Shetty, V., 2011. Mobile phones: emerging tools for

medical testing. In Internet Evolution.

www.internetevolution.com.

Shah, SG., Robinson, I., 2011. Patients' perspectives on

self-testing of oral anticoagulation therapy: content

analysis of patients' internet blogs. In BMC Health

Services Research. 3, 11:25.

Shortliffe, EH., 2010. Biomedical informatics in the

education of physicians. In Journal of the American

Medical Association. 304, 1227-1228.

Academic Informatics Programs, 2011. In American

Medical Informatics Association (AMIA) website.

www.amia.org/education/programs-and-courses.

AMIA 10x10 Courses, 2011. Training Health Care

Professionals to Serve as Local Informatics Leaders

and Champions. In American Medical Informatics

Association (AMIA) website.

www.amia.org/education/10x10-courses.

Kim, SN., Martinez, D., Cavedon, L., Yencken, L., 2011.

Automatic classification of sentences to support

evidence based medicine. In BMC Bioinformatics. 12,

Suppl 2:S5.

Korhonen, A., Silins, I., Sun, L., Stenius, U., 2009. The

first step in the development of text mining technology

for cancer risk assessment: identifying and organizing

scientific evidence in risk assessment literature. In

BMC Bioinformatics. 22, 10:303.

BUILDING TODAY THE FUTURE CADRE OF INFORMATICS-ORIENTED PHYSICIANS - Lessons from a New

Biomedical Informatics Curriculum for Medical Doctors

377