CS Spotlight Series: A Conversation with Professor Ophir Frieder

By Mariam Khan, Georgetown University

As part of our new CS Spotlight Series at Georgetown, we spoke with Professor Ophir Frieder, whose focus is on scalable information systems and health informatics. An inductee of the Florida Inventors Hall of Fame and a Fellow of the AAAS, ACM, AIMBE, IEEE, and NAI, Professor Frieder is on the faculty of the Department of Computer Science in the College of Arts and Sciences and the faculty of the Biostatistics, Bioinformatics, and Biomathematics in the Medical Center at Georgetown University.

What is your main research focus area and what inspired you to pursue it?

I’ve been involved with health informatics in one way or another for decades. I got involved as a grad student helping another grad student on work related to medical imaging. Over the years, I helped develop systems that dealt with the forerunner of electronic medical records, although it saw minimal interest at the time. Later, I worked on understanding modalities, but again, it was early days.

The real interest came around 2000 with diagnostics in collaboration with a urologist at Northwestern University. We developed a system to predict pathogens and treatments for urinary tract infections, which at the time was the number two reason women within the range of 16 to 65 years old went to doctors in the U.S. Given associated antibiotic resistance and that treatment guidelines were updated only once a year, we created a system that predicted which medication was likely to work—and it outperformed the guidelines.

We tried to get NIH funding, and they were interested. Unfortunately, after September 11, 2001, funding priorities shifted. So, the university filed a patent, and that marked the heyday of my work in medical informatics. I later expanded into different domains: mental health, drug prediction, COVID detection, and so on. My primary research area is scalable information retrieval, but I apply search techniques heavily in health informatics.

Was there a personal motivation behind your interest in this field?

I grew up in Buffalo, New York, a major center for cancer treatment and medical imaging. My father worked at the forefront of medical imaging with the Medical Imaging Processing Group, a leading center in computerized medicine. As a kid, I played with imaging systems at a time when most kids hadn’t even seen a computer. My high school offered a computer programming class—rare for the time—but only for the “weird” kids interested in this new thing. Seeing doctors work with technology made me realize early on how impactful computing could be in medicine.

What are some challenges you’ve faced in the field?

The biggest challenge is access to data. Medical data are heavily regulated, making them hard to obtain at scale, and rightfully so. Even when you get the data, the human body is inherently inconsistent—unlike mechanical measurements, biological measures vary widely. Normal or average metrics are often misleading.

Understanding complex medical data requires close collaboration with medical experts. It’s not enough to just analyze; you must interpret correctly, and that demands trust and effective collaboration. Also, regulatory barriers make it nearly impossible to develop diagnostic tools easily. In academia, my teams build tools that aid, filter, or alert— but not diagnose directly. Diagnostic tools are often classified as medical devices, and those require industrial-level ruggedization and compliance, which universities are not set up for.

You mentioned external collaborations. Could you tell us more?

Georgetown doesn’t have an engineering school—that’s a fact, not a judgment—so if you want to develop systems, you need external partnerships. For engineering, I collaborate with multiple universities, including Carnegie Mellon and Oregon State, to name a few. In terms of medical partners, I collaborate with doctors at Georgetown and other institutions like Cincinnati Children’s Hospital and Irvine hospitals. I’ve worked with startups and research labs worldwide, including in Italy. Some collaborations are pure research; others involve licensing agreements for Georgetown or consulting work.

What do you enjoy most about your work?

Seeing something change. Publishing papers or winning grants doesn’t move the needle by itself. I want my work to have real-world impact—to improve lives or, at minimum, to “do no harm.” Over the years, I’ve seen changes happen through collaborations, startups, and mentoring doctoral students who have made a difference in academia, policy, and industry.

Could you share an example of how you approach problem-solving?

It’s easier to define what’s wrong than what’s perfect. For example, improving visibility in degraded historical documents: instead of trying to “enhance” the good parts, we focused on removing the bad parts—like noise. It’s like sculpting: if you remove everything that’s not the statue, the statue appears.

Similarly, it’s easier to describe what isn’t beautiful than to define beauty. Overweight, missing features, disproportionate—those are easier to identify. If you eliminate enough negatives, the positive emerges. Many of the techniques I’ve used over the years rely on that principle: find and remove what’s wrong, and what’s left is better.

What advice would you give to students pursuing this field?

Never chase technology. By the time you graduate, today’s groundbreaking technology will either be widely understood or obsolete. Focus instead on identifying important problems and pitfalls.

When I wrote my bestselling book in 2004, Twitter and Facebook didn’t exist. Someone later asked why the book didn’t cover social media. I told them, “Because if I could predict the future before it happens, I would do much more than simply write a book about it.”

Technology changes fast. What stays valuable is your ability to identify meaningful, solvable problems. I tell my PhD students: if you already know what you’ll work on when you start, find another advisor. Research evolves. You must be inquisitive, willing to find problems, assess their feasibility, and adapt as the field changes.

Read more about Professor Ophir Frieder’s career achievements here.