Honeycomb-type graphic illustrating the program's four teams

Research

Researchers in the Pancreatic Cancer Early Detection Research Program are using multiple approaches to develop new tools for the early detection of pancreatic cancer, collaborating not only within our comprehensive investigative program but across all of Mayo Clinic as well.

We're building on decades of work that has made Mayo Clinic a highly regarded leader for advancing early detection of pancreatic cancer. Our program brings together experts who include physicians, epidemiologists, data scientists, laboratory technicians, bioinformaticians and clinical research staff.

For example, in partnership with researchers in the Department of Artificial Intelligence and Informatics and the Department of Radiology, we're leveraging a robust collection of well-characterized clinical and imaging data to develop artificial intelligence (AI)-based algorithms and combine the data with biomarkers to build a comprehensive approach to detecting pancreatic cancer early.

With four strategically designed teams, our innovative approach is addressing several unmet needs in the field of early detection:

  • Early detection biomarker team. This team's research focuses on optimizing and validating novel biomarkers using a blood-based approach to early detection. Future development includes a blood-based screening test for people who face a high lifetime risk of pancreatic cancer either because of genetic mutations or family history of pancreatic cancer, or both. Additional biospecimens (pancreatic juice, cyst fluid, stool and urine) are being collected from research participants to establish a bank of samples for future analysis.
  • Imaging research team. This team's current projects focus on developing artificial intelligence methods to facilitate semi-automated detection of early pancreatic ductal adenocarcinoma; to use a convolutional neural network trained for pancreas segmentation of CT and MRI images; to develop a deep learning algorithm capable of differentiating benign from neoplastic etiology; and to assess intrapancreatic fat on imaging as a risk factor for pancreatic cancer.
  • Natural language processing team. Work done by this team includes developing machine learning-based algorithms for identifying people at high risk of pancreatic cancer using a combination of risk prediction models and leveraging institutional expertise in pancreatic diseases, informatics and natural language processing, and text analytical methods. The team also is using natural language processing algorithms that aim to identify novel risk factors for pancreatic cancer.
  • High-risk pancreas care team. Led by program director Shounak Majumder, M.D., and supported by nursing and genetic counselors, with access to pancreas surgeons and oncologists when necessary, this team runs the High-Risk Pancreas Clinic for patients at increased risk of developing pancreatic cancer. A clinical translational research team partners with the clinic and offers patients participation in a clinical registry related to the early detection of pancreatic cancer.