Master's in Medical Informatics
Harvard-MIT Division of
Health Science and Technology (HST)

HST has established a master's degree program to train individuals who exhibit a deep understanding of both clinical medicine and informatics issues. The core of the program emphasizes fundamental issues of computer languages and systems, and the application of artificial intelligence methods to providing decision support in clinical information systems. A strong program of individual guidance allows students to tailor their educational programs to their individual interests.

The program is open only to students who have already achieved advanced training in the health sciences, e.g., doctoral degree in medicine, dentistry, nursing, veterinary medicine, clinical psychology, or Ph.D. in medically-relevant field such as physiology. Most students will be Fellows in one of the Boston-area Medical Informatics training programs, though the program may also be open to medical students and others. A strong background in computer science and information technology is not a prerequisite, but definitely a goal of the program.

We expect graduates to become leaders in the development and application of the field of medical informatics. Those trained at the Masters level may pursue several alternative career paths. Some may view their informatics training as an adjunct to their primary training in medicine, helping them be more effective users of information processing technologies in their medical and/or research careers. Others may use their familiarity with information processing to become involved in the application of informatics technologies in hospital information systems or other medically-related computing activities.

Doctoral training in medical informatics is available through the MIT Department of Electrical Engineering and Computer Science's graduate program in computer science ("Area II"), and is appropriate for the strongest students who have the desire to study long and deeply, and plan to pursue academic and research leadership roles in their careers.

Requirements

To qualify for a Master's degree, a student needs to complete 66 units of graduate credit plus a Master's thesis (normally 24 units). Of the 66 units, 42 must be Graduate-H units, which are awarded for "serious" classes with substantial prerequisites, and with homework, examinations and projects in some combination. The additional 24 units may be from less structured activities, such as units from supervised research, units received for acting as a teaching assistant, etc. In order for the SM degree to be specified as being in a particular field (e.g., medical informatics), 36 of the 42 Graduate-H units must be in that field. The thesis is intended to be a project that can be completed with half-time work for a semester. Despite this minimal specification, it normally takes a student a year of elapsed time to bring a Master's project to fruition.

The typical MIT class is 12 units. Hence, a plausible program of study for a student in the proposed program would consist of:

Most graduate students at MIT have responsibilities (e.g., as research or teaching assistants) that occupy about half their time, thus restricting them to taking no more than 27 units of classwork per semester (though they may earn additional units through their RA or TA). Most Fellows who are expected to participate in the Medical Informatics Master's program will also have other responsibilities at their clinical institutions; they must often also engage in clinical activities and are expected to carry on a productive stream of applied research. Therefore, although it is theoretically possible to complete the above program in one year, a two-year program is likely to be much more typical.

The typical student will take the following classes:

  1. The MIT introductory computer science class, 6.001, Structure and Interpretation of Computer Programs. This class is not suitable for graduate-H credit.
  2. HST947, Medical Artificial Intelligence, which is based on 6.034, the MIT undergraduate introductory class in Artificial Intelligence, supplemented by additional section meetings that focus on the medical application of AI techniques. Although 6.034 is an undergraduate class, this section is taught under a graduate number and is granted graduate-H credit.
  3. A comprehensive survey of clinical computing, 6.872/HST950. This is a graduate-H class.
  4. 6.891/HST951, Medical Decision Support, which examines the scientific principles and engineering approaches to building computer systems that assist in decision making.
  5. Elective class(es), depending on the interests and needs of the student. A long list of such classes is presented below:

The first three are header subjects that introduce different areas of computer science and engineering to our undergraduates. These classes do not get graduate-H credit but are intended to provide relevant background for those students whose educational program depends on material in these classes and who have not yet been exposed to that background.

6.033
Computer System Engineering; has as prerequisites (transitively, to include all their prerequisites) 6.004 (Computation Structures), 6.001, and 6.002 (Circuits and Electronics).
6.046
Introduction to Algorithms; has as prerequisites 6.001 and a discrete math class.
6.011
Introduction to Communication, Control and Signal Processing; has as prerequisites 6.001, 6.002 , 6.003 (Signals and Systems), and 6.041 (Probabilistic Systems Analysis).

More advanced (graduate level in computer science) classes that are probably relevant to the program include the following. (This list is meant to be suggestive, not exhaustive.) A student with particularly strong preparation in a certain topic may choose to take the graduate level class in place of a more introductory class listed above. Some of these classes will be accessible to students only if they have advance preparation in their domain or if they take preparatory classes leading up to these.

6.821 -- Programming Languages
6.823 -- Computer System Architecture
6.826 -- Principles of Computer Systems
6.835 -- Concurrent Systems for Artificial Intelligence
6.836 -- Embodied Intelligence
6.837 -- Computer Graphics
6.845 -- Parallel Processing: VLSI and Microarchitecture
6.846 -- Parallel Processing: Systems Architecture and Applications
6.847 -- Dataflow Architecture and Languages
6.848 -- Theory of Parallel and VLSI Computation
6.851 -- Theory of Algorithms
6.852 -- Distributed Algorithms
6.853 -- Computer Systems
6.854 -- Advanced Algorithms
6.858 -- Machine Learning
6.863 -- Natural Language and the Computer Representation of Knowledge
6.866 -- Machine Vision
6.868 -- The Society of Mind
6.871 -- Knowledge Based Application Systems
6.875 -- Cryptography and Cryptanalysis
6.231 -- Dynamic Programming and Stochastic Control
6.431 -- Applied Probability

Classes in the Sloan School of Management may be appropriate for some students:

15.065 -- Decision Analyis
15.136 -- Health Technology
15.138 -- Seminar on Pharmaceutical and Biotechnology Industry Management
15.141 -- Comparative Health Systems
15.563 -- Inventing the Organizations of the 21st Century
15.564 -- Information Technology I

Classes at the Harvard School of Public Health may also be appropriate for some students. Below is a small selection from the much more extensive HSPH catalog.

HPM204(d)
Research Synthesis and Meta-Analysis Applications in Public Health and Clinical Medicine
BIO200(ab)
Introduction to Statistical Methods
BIO 248(cd)
Advanced Statistical Computing
BIO 265(ab)
Nonlinear Repeated Measurement Models
Epi 201a
Principles of Epidemiology
HPM 234a
Managing in Health Organizations
HPM238(c)
The Strategic Use of Information Systems
in Health Care Delivery

Other programs at MIT also offer relevant classes. For example:

TPP 42J -- Ethical Problems in Advanced Engineering and Science
9.369 -- Computational Neuroscience
9.371 -- Advanced Seminar on Inference and Learning
9.520 -- Learning, Approximation, and Networks

Classes such as these cannot cover all of the topics that we believe to be important to a medical informatics student. However, when coupled with experience in a research laboratory, participation in optional seminar series, and mentoring by the senior staff in the program, students should receive a valuable and broad exposure beyond the specialization of the classes they attend.

The above proposal assumes (by choice of its required subjects) that most Masters candidates will be interested in those aspects of medical informatics concerning record keeping and decision making. This is consistent with the current focus of the Boston-area training programs, but is not the only type of student for whom a Masters in Medical Informatics makes sense. For example, a similar student who is particularly interested in image processing, radiological sciences, etc., should be able to adapt this program to fit his or her needs.

Admissions

Inquiries about admission to the HST Medical Informatics Master's program should be addressed to

Medical Informatics Training Program
ATTN: Robert A. Greenes, MD, PhD
Decision Systems Group, Brigham and Women's Hospital
75 Francis Street, Boston, MA 02115
(617) 732-6281 FAX: (617) 732-6317
email: greenes@harvard.edu