February 11, 2025

Spotlight: Martine De Cock’s privacy-preserving artificial intelligence research

The potential for artificial intelligence to solve some of the world’s biggest problems has always motivated Dr. Martine De Cock to work on developing and improving its utilization across multiple disciplines.

De Cock, a professor in the School of Engineering & Technology at University of Washington Tacoma, particularly believes in the power of privacy preserving AI in healthcare, which involves securely obtaining and utilizing patient data in order to improve current machine learning models aimed at diagnosing and treating a variety of diseases and illnesses.

“Working at the intersection of AI and healthcare is not easy because if you want to do it in a meaningful way, you have to collaborate through interdisciplinary research,” said De Cock. “You have to learn to speak the language of people who are in healthcare, and they have to understand you.”

De Cock began her academic career at Ghent University in Belgium, where she obtained a Masters of Science and Doctoral degree in computer science and later pursued research as a faculty member in the school’s computer science department.

De Cock came to the United States as a visiting scholar at UC Berkeley in 2003 and later pursued research in machine learning projects at UW Tacoma’s Center for Data Science. Her research led to a position as a visiting associate professor in 2008, where she focused on applied machine learning research that included healthcare data analytics.

However, De Cock faced administrative barriers and red tape to effectively obtain sensitive patient data for her machine learning research, which forced her to shift her research focus to social media data. This adversity inspired her to work on privacy preserving AI that could safely use patient data to train and use AI models in healthcare.

“Deep down, I felt frustrated by that because I think the most important problems in the world for us to solve, and that machine learning can solve, are not necessarily about social media,” said De Cock. “There’s a lot that AI can do for healthcare, but it’s difficult to get started when you don’t have access to the data.”

De Cock’s privacy-preserving AI research has involved close collaboration with faculty in UW Tacoma’s School of Engineering and Technology. For nearly a decade, she conducted research in privacy-preserving machine learning with Dr. Anderson Nascimento, an expert in information theory and privacy. Last year, she teamed up with Dr. Paulo Barreto, an expert in cryptography. Their proposal on synthetic data generation recently earned them one of 35 National Artificial Intelligence Resource (NAIRR) Pilot First Round Access Awards granted nationwide by the National Science Foundation (NSF).

De Cock has also worked closely with Weichao Yuwen, an associate professor in UW Tacoma’s School of Nursing and Healthcare Leadership. The two received a capacity building grant from the NIH about two years ago, which allowed them to investigate the intersection of health equity and AI. Their research funding created UW Tacoma’s Responsible Health AI Lab, which fosters an interdisciplinary space for students in UW Tacoma’s School of Engineering & Technology and the School of Nursing & Healthcare Leadership to create innovative applications of artificial intelligence that address a wide range of challenges related to disease prevention, diagnosis and treatment, and self- and family management of chronic health conditions.

“My whole research trajectory has been very much influenced by the people who came on my path, and then I started working together with them for every project I did,” said De Cock. “Every new turn that I took, there’s always been an influential person who I could collaborate with and who’s had a big influence on me and my work.”

De Cock is currently a principal investigator for a Population Health Initiative Tier 1 pilot grant, where she, Yuwen and Dr. Sarah Iribarren at UW Seattle’s School of Nursing, are working to develop a prototype AI chatbot that can assist health providers in quickly and effectively supporting patients with tuberculosis. She attributed her recent successes to the UW Population Health Initiative, which connected her with faculty in nursing and took her research on privacy preserving AI for health to the next level.

“It’s not only the research funding, the money part that is interesting but it’s the other opportunities that are created through the Population Health Initiative,” said De Cock. “The PHI inspired me to talk with faculty in nursing. I was very fortunate to find a wonderful collaborator in Dr. Weichao Yuwen. Fast forward two years, and we have a new responsible health AI lab that brings together undergraduate and graduate students across our disciplines. For an emerging research institute like UW Tacoma, with a young Ph.D. program in computer science, that is very significant.”

As De Cock continues her research on privacy preserving AI, she believes that the Population Health Initiative holds the key to advancing AI use in healthcare through both funding opportunities and meaningful interdisciplinary collaboration.

“I’m really grateful for the Initiative to bring people together who are working in AI and health, and not just for me, but for my collaborators and my students at UW Tacoma,” said De Cock. “It’s a way for them to connect with other researchers and build their network within UW and beyond. It’s not just about the money, it’s about bringing people together.”