Applied Machine Intelligence

The Applied Machine Intelligence Program in the Mayo Clinic Kern Center for the Science of Health Care Delivery develops innovative practice solutions using advanced computational and machine learning methods. The team optimizes the capacity of Mayo Clinic's hospital and clinical service and increases Mayo's ability to care for patients with serious conditions who need complex care. At the same time, the program team prioritizes workforce well-being in every project across diverse practice areas.

In the Applied Machine Intelligence Program, experts use visualization methods and computational data from biological and clinical knowledge domains to develop new theories and algorithms for machine intelligence. Team members engage in research across all facets of machine learning, including:

  • Classical machine learning algorithms.
  • Deep graph learning.
  • Deep learning.
  • Immersive intelligence.
  • Large language models.
  • Natural language learning.
  • Probabilistic inference.

Program experts explore the feasibility, efficiency and effectiveness of translating these technologies into clinical practice.

Solutions developed within the Applied Machine Intelligence Program support Mayo Clinic's multidisciplinary medical care by facilitating point-of-care solutions personalized for each patient — no matter how complex the individual's care needs. The program creates these solutions by drawing on knowledge and data from hundreds of millions of patient characteristics and care episodes.

The program develops intelligent machines that produce solutions — and then improve on those solutions — without direct human intervention. Team members aim to integrate their advanced artificial intelligence solutions into clinical practice, thereby transforming healthcare for today and tomorrow.

Focus areas

The Applied Machine Intelligence Program is focused on:

  • Access. Enhancing access to information by delivering relevant, just-in-time knowledge to clinicians and patients.
  • Automation. Automating clinical workflows and administrative tasks, such as pre-authorizing insurance and submitting claims after care.
  • Development. Developing new applications for patient-clinician communication, improving existing tools and eliminating barriers to equitable healthcare.
  • Efficiency. Improving the efficiency and generalizability of clinical trials and reducing disparities in access and participation.
  • Safety. Identifying and characterizing patient safety hazards and providing point-of-care recommendations for alternative care pathways that could lead to better outcomes and patient experiences.

Affiliations

The Applied Machine Intelligence Program works with partners across Mayo's clinical practice. Recent collaborators include:

  • Department of Obstetrics and Gynecology
  • Department of Surgery
  • Division of Endocrinology, Diabetes, Metabolism, and Nutrition
  • Division of Medical Oncology

Projects

The program has multiple collaborative projects in development. Research aimed at advancing "intelligent" technology is designated with an "i" at the beginning of the project name. Projects include:

  • iMidWife. An end-to-end artificial intelligence system for labor and delivery that supports obstetrics staff and patients for a safer and more comfortable birth experience.
  • iPCa. Use of extended reality — which can include virtual, augmented or mixed reality technologies — for risk assessment and psychosocial support among patients with prostate cancer.
  • iRareGene. Large language models for identifying patients with rare genetic disorders, which could shorten their diagnostic journeys and lead to more informed decision-making.
  • iRemoteCare. Artificial intelligence-based trajectory models for clinical decompensation in intensive care units and among patients receiving remote monitoring care.
  • LLaMPS. A new large language model for patient safety along the entire continuum of care.

Contact

Che G. Ngufor, Ph.D.