As a student, you were interested in both medicine and mathematics, how did you pursue the two?

At school, mathematics was my stronger subject and the one I enjoyed the most. I wanted to find a way to use mathematics in health care, initially with the idea of pursuing bioengineering. But growing up in Brazil I had to pursue a medical or an engineering degree and then find ways to combine my interests. I decided it would be easier to start with a medical degree and build on that.

How did you navigate career opportunities in the emerging field of biomedical informatics?

You must be a risk taker. You can't possibly be too cautious when pursuing a career that is not yet seen as mainstream. My self-confidence helped me overcome the barriers created by people's inability to understand the potential of an emerging field. Twenty years ago, the United States had the right combination of programmes available and was open to interdisciplinary careers, so I came to Stanford University to get a PhD in medical information science.

Will the new graduate programme at the Division of Biomedical Informatics differ from other similar programmes?

Yes. Several programmes funded by the US National Institutes of Health do bioinformatics or medical informatics. But only a few of those do both, and even fewer provide trainees with access to clinical information systems. The San Diego programme will tightly combine bioinformatics and medical informatics while making use of data from the medical-centre hospitals, which will be available for our students and trainees as their laboratory.

What drives your interest in finding patterns in clinical data?

We are past the age when the experience of one clinician can lead to significant medical advances. We need to find practical ways to access all the data from a larger collection of patients being treated. If we can extract patterns that tell us what we are doing right and what we can do better in the clinic, it will help patients and health-care providers. For example, data registries exist for certain procedures, such as interventional cardiology, and these may be helpful to the clinician assessing the complications associated with a new device.

What is the biggest challenge in biomedical informatics?

Achieving the same status as an established science. That status will attract more young scientists who will shape the future of biomedical informatics. Well-established programmes, developed since the 1980s, have paved the way for the dissemination of biomedical informatics around the world. It will not take long for every major academic institution to develop its own department in this field.