2023
11:00 AM - 12:00 PM
Monday Mon
Location
Stanford University School of Medicine
291 Campus DrStanford, CA 94305
Medical Physics Seminar - Charles Mayo
Local and Global fronts in Combining Big Data & AI to Improve Cancer Care MROAR, OORO and Standardizations
Time:
12:00pm – 1:00pm Seminar & Discussion
Location:
Zoom Webinar
Webinar Registration:
https://stanford.zoom.us/webinar/register/WN_j1zCOA3wTUaOW1k9ue6zHQ
Check your email for the Zoom webinar link after you have registered
Speaker
Charles Mayo, PhD, FAAPM, FASTRO, Professor Director of Radiation Oncology Informatics and Analytics
Dr. Mayo did their doctoral work at the University of Massachusetts at Amherst where they developed an optical technique to investigate antibody - adsorption kinetics using surface plasmon resonance as a probe. Their post-doctoral work was in proton radiation therapy at the Harvard Cyclotron Laboratory where they focused on particle beam line calibration including discovery of characterized edge scattering effects and design of database applications for routine calibration. Dr. Mayo then expanded their base toward external beam photon and brachy therapy. Their enthusiasm for combining technology, clinical practice and standardization has its roots in the work that I did for the Quantitative Analysis of Normal Tissue Effects in the Clinic (QUANTEC) initiative, working on both the brainstem and the optic nerve/optic chiasm papers. Overcoming the full scope of barriers to using the vast amount of potentially useful information in our electronic systems, so that findings based on thousands instead of tens of patients could become routine, became the driving force behind their research interests.
Dr. Mayo’s primary research interest centers on constructing standardized large scale data bases from routine practice data for use in integrating artificial intelligence driven modeling into clinical care. As Director of Informatics and Analytics in the Department of Radiation Oncology, they led their team in creation of their platform, named the Michigan Radiation Oncology Analytics Resource System (MROAR), that aggregates, integrates, and harmonizes data from the several commercial data systems used to treat radiation oncology patients into a single, easy to use platform. Their work with that system includes developing novel methods and visualizations combining statistical and machine learning methods to construct outcomes models from comprehensive, large scale, “real-world” data to detail actionable clinical thresholds for the purpose of reducing incidence of radiation dose related toxicities and other undesirable outcomes such as emergency room visits. In addition, Dr. Mayo works on development of software applications to automate treatment planning incorporating histories of plans from previously treated plans collected in MROAR. Leading large, multi-disciplinary, multi-institutional stakeholder groups in developing standardizations supporting interoperable data exchange is a key component of their primary interest to support creating multi-institutional databases. They led the first multi-professional society led effort to establish a common nomenclature for dosimetric data in radiation oncology as chair of Task Group 263 of the American Association of Physicists in Medicine (AAPM) and chair of the subsequent Radiation Therapy Standardizations (SC-263) committee.
As vice chair of AAPM’s Data Science Committee and chair of the Big Data Subcommittee (BDSC), Dr. Mayo is leading another large, multi-disciplinary, multi-society effort with the American Society for Radiation Oncology (ASTRO) to define a standardized operational ontology for radiation oncology (OORO). Working with AAPM and ASTRO, they are a member of the radiation therapy (RTTD) use case team on mCode and CodeX projects with Miter Corporation and HL7 in creation and implementation of HL7-FHIR tags supporting vendor neutral transmission of radiation oncology specific data elements.
Local and Global fronts in Combining Big Data & AI to Improve Cancer Care MROAR, OORO and Standardizations
Radiation Oncology is a uniquely data driven specialty in medicine. Creating more data and AI driven clinical practices requires combining efforts along several fronts 1) integration of data culture into clinical practice 2) informatics and infrastructure, 3) standardizations and 4) development of explainable AI models and 6) integration into clinical practice. This presentation will highlight our combinations of local and multi-institutional efforts to create more data centric practices in radiation oncology.
A video recording of the presentation will be available upon request. Please email jmakim@stanford.edu for requests.