This story is adapted from an article by Becky Sun that ran in the January issue of The Connector, a newsletter of the Harvard-MIT Division of Health Sciences and Technology.
Even as a young girl in Brazil, Lucila Ohno-Machado knew what she wanted to do when she grew up. "I've always wanted to work with computers in medicine," she said.
Today the professor with joint appointments at MIT and Harvard Medical School is working on computer models that will help plan treatment for diseases using the patient's medical history and diagnosis.
Ohno-Machado has several ongoing research projects in medical informatics, a field that uses biomedical information and data to solve problems and make decisions in health care. For example, she is using algorithms to construct diagnostic and prognostic models of diseases.
"Diagnoses are things that [doctors] do well since that is taught more in medical school. But people don't do as well with prognoses," she said. "Recent developments in functional genomics suggest that it may be possible to do even more precise diagnoses if gene expression is taken into account. Since the new methodology is heavily dependent on computation, we can anticipate a critical need for computers to help clinicians make diagnoses and prognoses."
Ohno-Machado also is working to automate the process of selecting patients for clinical trials of new therapies. Such trials are an effective way to deal with advanced breast cancer, but there are hundreds of trials at any one time with very specific requirements. Ohno-Machado's goal is to design a computer program that can sort through the myriad criteria using patient data to find the most appropriate trials.
Ohno-Machado teaches medical informatics to health care providers and computer scientists to bridge the gap between the two fields. Last year she received a Taplin award from HST to develop and build a biomedical informatics curriculum. Her approach stresses the need to be responsive to the expectations and anxiety of cancer patients. She is an ardent believer in the need for "bench and bedside" methods to communicate and collaborate with patients.
Besides her research and teaching, Ohno-Machado, who also is associate director of the Decisions Systems Group at Brigham and Women's Hospital, has a strong desire to give back to her native country. In 1999 she received a grant from the National Institutes of Health's Fogarty International Center to do medical informatics training in developing countries.
The first part of the training is to send specialists from the United States to Brazil to teach short courses. (HST faculty members Robert Greenes and Peter Szolovits have taught courses there.) The second part of the grant brings Brazilian researchers to the United States for training.
This winter, in response to an NIH request for applications for international collaboration in genetics training, Ohno-Machado worked with Dr. Richard Maas and Dr. George Church, both affiliated faculty of HST, to plan a rigorous post-doctoral training program for young Brazilian researchers.
"Brazilians have actively participated in the Human Genome Project and similar research, such as sequencing the genome of certain plants and infectious diseases of national importance," she said. The Brazilian research community in genomics and bioinformatics, although extremely talented, lacks active support for collaboration with leading international centers of excellence in genomics and bioinformatics. The proposed program tries to fill this gap.
In addition to her research, teaching and advising responsibilities, Ohno-Machado is mother to three boys ages 6, 4 and 2. She had been able to juggle her work and home life successfully until recently, but "the third one unbalanced me," she said half jokingly. She and her husband divide the child care, with him staying home for now.
A version of this article appeared in MIT Tech Talk on April 10, 2002.