New funded projects from Joint Center for Computational Oncology integrate predictive digital twins to improve patient outcomes

UT MD Anderson News Release October 16, 2026

The University of Texas MD Anderson Cancer Center, the Oden Institute for Computational Engineering and Sciences and the Texas Advanced Computing Center (TACC) at The University of Texas at Austin have announced seed funding for two new cancer research projects as part of the Joint Center for Computational Oncology (JCCO).

Both of this year's projects aim to create advanced predictive models known as 'digital twins,' which build digital models of real-world systems, such as patients’ tumors. Unlike other models, digital twins are continuously informed by real-world data, allowing researchers to simulate different treatment options that can help inform clinicians about those most likely to benefit a specific patient.

The projects will pair expert clinicians and scientists from UT MD Anderson with advanced computational scientists from the Oden Institute.

Since the JCCO program began in 2020, 29 projects have been funded, each of which bring together UT Austin and UT MD Anderson researchers. In addition to funding, project support includes 12,500 core hours on TACC systems to support large-scale computation.

Twenty years ago, the Oden Institute’s first collaboration with UT MD Anderson began when John Tinsley Oden, Ph.D., founder of the Oden Institute, initiated research with John Hazle, Ph.D., chair of Imaging Physics at UT MD Anderson, on predictive modeling of laser-ablative therapy for prostate cancer. This area of research became a passion for Oden, and from that early relationship, the Joint Center for Computational Oncology was born. Hazle now co-leads the effort with Tom Yankeelov, Ph.D., director of the Oden Institute’s Center for Computational Oncology.

The collaborative initiative between the institutions continues to accelerate breakthroughs in oncology by combining the Oden Institute’s expertise in computational science, leadership in clinical oncology and data science from UT MD Anderson’s Institute for Data Science in Oncology (IDSO), and TACC’s advanced high-performance computing capabilities.

“Now in its seventh round, the collaboration continues to bring clinicians and computational scientists together around some of the most pressing problems in cancer care,” Yankeelov said. “This year's projects build on lessons from earlier rounds — pushing into new disease areas while staying focused on solutions that can move quickly from the lab into clinical application.”

“These collaborations are more essential than ever in advancing cancer research,” Hazle said. “Both projects exemplify the interdisciplinary approach needed to build models that create tangible benefits for patients. These seed funds help form the teams that scale and go on to secure additional funding to drive the innovations to the clinic.”

2026-2027 Funded Research Projects

A Multimodal Digital Twin Framework for Predicting Cervical Cancer Response to Chemoradiation

Led by Ernesto A. B. F. Lima, Ph.D., research scientist at the Oden Institute and TACC, and Priya Bhosale, M.D., professor of Abdominal Imaging at UT MD Anderson, this project aims to build a digital twin of cervical cancer for each patient. The model will combine imaging with clinical and treatment information to track how tumors change and predict how they may respond to treatment.

Contrast-Sparing Imaging Twins for Longitudinal PRRT Monitoring of pNET Liver Metastases

Led by Chengyue Wu, Ph.D., assistant professor of Imaging Physics at UT MD Anderson, and Hairong Wang, Ph.D., assistant professor of Operations Research and Industrial Engineering (ORIE) and affiliated faculty member at the Oden Institute, this project will develop imaging twins to monitor liver metastases from pancreatic neuroendocrine tumors during peptide receptor radionuclide therapy. The team will analyze sets of non-contrast and contrast-enhanced CT scans acquired before and after treatment to develop a model that generates virtual contrast-enhanced CT images from non-contrast scans. Where contrast scans require contrasting agents to be delivered to patients, either via intravenous needle or catheter injection, this model will generate the same images from non-invasive imaging. These generated images will complement PET and SPECT measurements, providing a more complete view of tumor changes throughout treatment.