The Clinical Decision Making Group (CDM, aka MEDG) of the MIT Computer Science and Artificial Intelligence Laboratory provides a setting for research training in Bio-Medical Informatics, and is currently the primary point of contact with MIT for Fellows in the Boston-area Informatics Training Program, centered at Harvard Medical School. We focus on research in the application of computing methods to important problems arising in health care. The group's home page lists some of our current and past research foci. Prof. Szolovits also teaches the class 6.872/HST.950, Biomedical Computing, which provides a graduate-level introduction to this field. During the school semesters, we also hold a weekly research seminar series. (If you are interested in getting email notices of this, please contact Fern DeOliveira.)
Training opportunities exist for the following categories of students working in the group. Follow the hyperlinks for additional information on requirements and admission information for each category.
If your situation does not match any of the above categories, then we probably do not have a training program appropriate for you. More general information is available on the EECS Department degree programs for undergraduates and graduate students. In addition, there are other research groups within the Laboratory for Computer Science and the Artificial Intelligence Laboratory that may provide an appropriate context for some biomedical informatics students: the Medical Vision Group, the Computational Biology Group, the Computational Genomics Group, the Computation & Biology Group, the Evolutionary Design and Optimization Group, and groups led by Dr. Seneff and Prof. Guttag. We estmate that nearly half of the faculty in EECS are engaged in some form of biomedical research, so there are many opportunities across EECS and CSAIL to find research supervision.
Medical Informatics (MI) is the study of information processing as it is used in healthcare. It might have been called medical computing, but the French-derived term informatique is more commonly used internationally and probably conveys a broader set of concerns, including the uses and flows of information that may have little to do with computers. Like many engineering fields, MI has scientific aspects that focus on the description, modeling and interpretation of how information is actually generated, disseminated and used, and underlying constraints or natural laws that govern these activities. MI is also deeply concerned with design of appropriate medical information processing systems, with tradeoffs in their implementation, and with ways to evaluate their effectiveness.
Some have suggested health informatics as a better, broader term, meant to encompass aspects of health care that are not traditionally the focus of medicine, such as preventive care, nutrition, patient education, epidemiology, etc. Related terms include bioinformatics, which is the study of information processing in biological sciences. Opinion currently varies on whether bioinformatics is part of medical informatics, or--if it forms a distinct discipline--how it relates. Most expect that progress in understanding the molecular basis of disease will bring these fields closer together, if not to merger. Telemedicine (or the recent European coinage telematique) focuses on one aspect of MI, access to and use of medical information at a distance.
At MIT, in line with our traditions of institutional flexibility, we have no official organization that does medical informatics, but a number of small foci around the research and teaching interests of faculty in different Departments and Laboratories.
Graduate education in computer science at MIT is conducted in Area II of the Department of Electrical Engineering and Computer Science (EECS). The graduate program awards master's degrees (S.M., or M.Eng. for students enrolled in a joint Bachelor's/Master's five-year program), and doctoral degrees (Ph.D. or Sc.D.). Currently, a master's degree is a prerequisite for pursuing doctoral studies, though students with other advanced graduate degrees (e.g., an M.D.) may petition to bypass this requirement. The Department does not encourage students who already hold a Ph.D. in another technical field to pursue further doctoral studies.
Through this program, there is no specific diploma in Medical Informatics, but students may specialize in this field through their selection of classes and thesis topics. Admission to the graduate program is very competitive, and does not distinguish between students whose initial plan is to complete only a master's degree and those who initially intend to pursue a doctorate. (We have found that people often change their plans.)
More information about admissions, financial aid, and admission applications may be obtained from the EECS Graduate Admissions Office. Admission in EECS is decided by a faculty committee acting for the department, and not by individual faculty members. Therefore, communicating with individual members of the faculty does not improve your chances of admission, contrary to what is apparently common practice at some non-US universities. Once a student has been offered admission, communication with faculty leaders of research groups is an excellent idea, to help decide whether MIT is in fact the place you wish to study, to plan your educational and research program, and to investigate the possibility of funding as a Research Assistant.
Beginning in the mid-1990's, EECS has offered a five-year joint bachelor's/master's program to our stronger undergraduate students. In the fifth year of this program, such students are in fact graduate students and act very much as other master's students. The M.Eng. program is open only to MIT undergraduates, and admission is normally granted to the majority of students in good undergraduate academic standing. M.Eng. students (unlike S.M. students) must apply separately for admission into the Ph.D. program if they wish to pursue doctoral studies. Financial aid is not assured, but M.Eng. students may be eligible for Teaching Assistanceships from the Department or Research Assistanceships from various sponsored research projects. As with other graduate programs at MIT, students specialize in medical informatics by choosing appropriate classes and by selecting a thesis topic in this area. Students wanting to find thesis topics in medical informatics should examine our group's research program and then contact us by email.
MIT encourages its undergraduate students to participate in research projects under the supervision of faculty, staff and senior graduate students. Many students participate in the Undergraduate Research Opportunity Program (UROP), which allows them to work on research either for academic credit or pay. Other opportunities for undergraduate research arise via the Undergraduate Advanced Project (UAP) requirement for graduation (which is typically fulfilled for M.Eng. students by writing their M.Eng. thesis proposal), and various special project courses listed in the MIT catalog.
CDM has had a very active and successful history of undergraduate research, and we encourage undergraduate students with an interest in medical computing and medical decision making to seek out and participate in our projects. Although projects (and funding) are not always available, we solicit expressions of interest and curriculum vitae by email. If you are seeking an opportunity to engage in undergraduate research, please refer to our group's home page to see the range of our current research activities and to determine whether one or another of our ongoing projects most interests you.
In the past, the mulit-institutional Informatics Training Program in the Boston area allowed students to pursue master's theses through HST, but that program moved in 2010 to become a graduate program of Harvard Medical School. The Fellows in that program continue to have the opportunity to pursue research projects with many institutions around Boston, including our group.
Students in other programs occasionally work with us. For example, some medical students at Harvard Medical
School do thesis projects focusing on medical applications of computing techniques. Graduate students in HST's
MEMP sometimes choose to do their doctoral research in our group. Finally, we occasionally have visiting scholars
working with us in the group, on projects of mutual interest. These types of students are rare because it is more
common for Harvard medical students and HST graduate students who work in this area to be simultaneously graduate
students in the EECS graduate program. Visitors who are students are generally discouraged, because MIT has no
provisions for visiting graduate students. Those who believe they would benefit from one of these more unusual
connections with our group should get in touch on an individual basis.
Last modified: 10/30/2011