image: Raffaella De Vita's lab group gathers to discuss research findings, with (from left) Lillian Fantuzzo, Raffaella De Vita, Clara Giménez, Mostafa Zakeri, Xiaomeng Wang, and Erika Hobson.
Credit: Photo by Ritika Singh for Virginia Tech.
The National Institutes of Health (NIH) has funded a Virginia Tech–led research project to better understand and predict women’s health outcomes by investigating how hormones shape the structure and function of vaginal tissue across the lifespan. The award for up to $9.26 million is part of a growing initiative by the NIH Office of Research on Women’s Health.
The researchers will use mathematical and computational modeling and machine learning to examine how vaginal tissue changes during pregnancy, childbirth and postpartum recovery, aging, and menopause. Ultimately the work could help better identify women’s health risks and develop strategies for prevention and treatment.
The large, multi-institutional team is led by Virginia Tech researcher Raffaella De Vita, professor in the Department of Mechanical Engineering, who has spent years studying vaginal and pelvic tissue biomechanics.
“What excites me most about this opportunity is the chance to ask a much bigger question: If we had all the data in the world, could we build a model that predicts how hormonal changes reshape vaginal tissue over time?” De Vita said. “I do not want to simply apply the computational tools we already have. I want the biological questions to push us to develop new mathematical, computational, and AI-enabled approaches — and then use those approaches to generate knowledge about women’s health that we could not obtain from experiments or data alone.”
Why does vaginal tissue change over time?
A woman’s hormonal environment changes substantially throughout life. Reproductive maturity, pregnancy, postpartum recovery, aging, and menopause bring changes not only in hormone levels but also in the composition, structure, and mechanical behavior of vaginal tissue.
But these changes and their outcomes vary widely among women. Pregnancy and childbirth can have very different long-term effects, as can the timing and experience of menopause. Understanding what drives that variability is a major challenge in women’s health, requiring expertise from engineering, mathematics, data science, reproductive biology, and clinical medicine.
How can AI-enabled research improve women's health?
The multidisciplinary team of researchers will integrate existing experimental, clinical, and population-level data to develop a new computational framework for predicting how vaginal tissue changes in response to hormones over time.
They’ll combine artificial intelligence and machine learning with mathematical and computational models to connect diverse data sets and identify relationships between hormone dynamics and changes in tissue microstructure and mechanics.
Uncertainty quantification will allow the team to determine how biological variability and limitations in the available data affect the reliability of their predictions. Then they'll test and refine their computational predictions against available experimental and clinical data, published findings, and other relevant observations.
“For a clinician, the ability to better understand how and why tissues change across different stages of a woman’s life could fundamentally change how we think about women’s health,” said collaborator Marianna Alperin, professor and vice chair for research in the Department of Obstetrics, Gynecology, and Reproductive Sciences at the University of California San Diego and a practicing urogynecologist. “The long-term goal is to turn that understanding into better ways to anticipate problems, prevent them when possible, and improve treatment.”
What experts will contribute to the research?
In addition to De Vita, contributors to the project include:
- Pinar Acar, associate professor in the Department of Mechanical Engineering at Virginia Tech
Acar works in uncertainty quantification and machine learning, developing data-driven methods for predictive modeling under uncertainty. - Marianna Alperin, professor and vice chair for research at the University of California San Diego
Alperin is a physician-scientist specializing in urogynecology and reconstructive pelvic surgery. Her research seeks to understand the mechanisms underlying pelvic floor disorders and translate scientific discoveries into improved strategies for women’s health. - Stanca Ciupe, professor in the Department of Mathematics at Virginia Tech
Ciupe develops mathematical models of biological systems, with expertise in describing how complex biological processes evolve dynamically over time. - Kathleen Connell, professor at the University of Colorado Anschutz School of Medicine
Connell is a physician-scientist and board-certified specialist in urogynecology and reconstructive pelvic surgery whose clinical and research expertise connects fundamental research with women’s pelvic health. - Justin Dubik, instructor and research collaborator in the Department of Mechanical Engineering at Virginia Tech
Dubik conducted research in De Vita’s laboratory before joining Virginia Tech’s instructional faculty and has contributed to experimental and computational studies of vaginal tissue biomechanics. - Traian Iliescu, professor in the Department of Mathematics at Virginia Tech
Iliescu brings decades of experience developing mathematical and computational approaches for complex systems, including methods that make large computational problems tractable. - Joshua Johnson, associate professor at the University of Colorado Anschutz School of Medicine
Johnson studies female reproductive biology, with expertise in ovarian function, reproductive aging, and the cellular and molecular processes that regulate reproductive health. - Justin Krometis, research associate professor in the Virginia Tech National Security Institute and affiliate faculty in the Department of Mathematics at Virginia Tech
Krometis brings expertise in parameter estimation and uncertainty quantification, helping researchers determine how variability and uncertainty in data and model parameters influence computational predictions. - Sean Lawley, professor at the University of Utah
Lawley develops mathematical approaches to biological and physiological problems, including research related to reproductive aging and menopause.
Together, the team brings complementary perspectives to a challenge no single discipline can address: building a more predictive understanding of how hormones shape vaginal tissue health across a woman’s lifespan.
NIH award number: OT2OD042813