The Society for Imaging Informatics in Medicine (SIIM) is planning its Conference on Machine Intelligence in Medical Imaging, SIIM CMIMI23, co-sponsored by American Association of Physicists in Medicine (AAPM), to be held in-person October 1-2, 2023 at Johns Hopkins in Baltimore, MD. Image courtesy: SIIM
August 31, 2023 — The Society for Imaging Informatics in Medicine (SIIM) is planning for participants to return for in-person education and networking during its upcoming Conference on Machine Intelligence in Medical Imaging, SIIM CMIMI23. Co-sponsored by American Association of Physicists in Medicine (AAPM), the event will take place October 1-2, 2023 at Johns Hopkins in Baltimore, MD. Early registration ends September 15 for the 2-day meeting. An overview of key program elements, the highly-regarded conference planners and participants, as well as session objectives follows.
Topics to be covered include Large Language Models, Generative AI, and Emerging Technologies. Scientific Abstracts presentations will focus on key areas: Clinical Applications; Intelligent Imaging; Data Sets, Emerging Technologies and Natural Language Processing (NLP) Models. Nearly 30 scientific posters will also be presented.
The conference is intended to help participants interact one-on-one with some of AI's most respected thought leaders from academia, clinic, and industry, while playing a role in advancing machine learning-based applications to the real world. Organizers for the 2-day conference also reinforced the wide-ranging opportunities for attendees to connect with peers from various areas — researchers, clinicians and students engaged in scientific and clinical research, as well as computer scientists, engineers and developers representing technology companies, and the startup community working on innovative solutions.
Conference Co-Chairs for CMIMI23 are:
Katherine P. Andriole, PhD, FSIIM, Associate Professor of Radiology, BWH, Harvard Medical School and Director of Research Strategy & Operations, Massachusetts General Hospital Brigham Center for Clinical Data Science
Jeffrey H. Siewerdsen, PhD, FAAPM, FAIMBE, Professor, Department of Imaging Physics and Director, Surgical Data Science Program, Institute for Data Science in Oncology at the University of Texas MD Anderson Cancer Center
Planners have noted that CMIMI23 participants will have a unique opportunity to engage in thought-provoking questions and conversations with other leading AI researchers, as well as the chance to meet top industry influencers in a smaller setting ideal for making connections.
The event will kick off with an Opening Keynote from Woojin Kim, Chief Medical Information Officer at Rad AI. A SIIM summary of his presentation, “Expanding Horizons: An In-depth Exploration of Generative AI in Medical Imaging," which will take place Sunday, October 1, from 8:00-9:00 a.m. ET.
Conference planners report that the keynote session will provide a comprehensive overview of the role and impact of generative AI in medical imaging, starting with exploring the immense potential that GenAI holds by looking at foundation models beyond ChatGPT. Kim will highlight cutting-edge research in large language models (LLMs) and their broad applications in medical imaging. Sharing insights on the untapped possibilities of GenAI beyond LLMs, the session will also focus on the challenges of GenAI technology, including its application in medical imaging, to foster a balanced understanding of the benefits and risks associated with this technology. SIIM CMIMI planners also noted that for the entrepreneurs in the audience, we will also explore the unique business challenges related to creating and deploying GenAI applications.
A highlight of the the SIIM CMIMI23 “Town Hall: Synthesizing Diagnostic Imaging Data Scientists,” to be moderated by Katherine Andriole, PhD, Massachusetts General Brigham, also a Conference Co-Chair, which will include five panelists. They include: Peter D. Chang, MD, University of California, Irvine; Bradley J. Erickson, MD, PhD, CIIP, FSIIM, Mayo Clinic; Elizabeth A. Krupinski, PhD, FSIIM, Emory University; Sharmila Majumdar, PhD, University of California, San Francisco; and Paul H. Yi, MD, MS, University of Maryland.
Planners note that This Town Hall session, set for 1:14-2:45 on Sunday, day one of the conference, will feature representatives from several successful AI Centers/Programs who have made major contributions to AI literature in diagnostic imaging. Each panelist will provide an overview of their program noting organizational structure, funding sources, recruitment, educational activities, ongoing projects, opportunities for collaboration, and keys to success.
The objectives for this special session, which will be followed by a panel discussion and Q&A with the audience, were identified as follows:
1) Review how various organizations are structuring and funding centers and/or activities in imaging AI; 2) Identify potential opportunities for project collaboration with existing centers; and 3) Discuss example educational activities, recruiting strategies and keys to success
Other sessions on the first day of SIIM’s CMIMI23 include:
“The Do’s and Don'ts of Publishing Machine Learning Manuscripts in the Journal of Digital Imaging” with Elizabeth A. Krupinski, PhD, FSIIM, Emory University, Editor in Chief, Journal of Digital Imaging (JDI)
“Choosing the Right Platform for Your AI Applications - A Vendor Panel Discussion”
Networking Reception, which will be held from 6:30-7:30 p.m.
Programs scheduled for Monday, October 2, highlighted on the SIIM CMIMI Conference online overview, are summarized below.
An "AAPM-SIIM Symposium on Machine Intelligence in Medical Imaging: A Medical Physics Perspective" will feature Karen Drukker, PhD, MBA, FAAPM, FSPIE, University of Chicago and Xun Jia, PhD, DABR, FAAPM.
Scientific Abstract Presentations will focus on: Generative AI & General Applications in NLP; Toolkits and Machine Learning Algorithms; and Clinical Applications, with sessions starting at 8:45 a.m., 10:30 a.m. and 1:00 p.m., respectively.
Prior to closing remarks on day two, a special session will be presented from 2:15 p.m. to 3:15 p.m. ET in the Turner Auditorium. “Regulatory Science for AI-enabled Devices in Medical Imaging: A Perspective from the AI/ML Regulatory Science Research Program at the Office of Science and Engineering Laboratories (OSEL) at the FDA” will include two speakers.
Victor Garcia, MD, Staff Fellow, Division of Imaging, Diagnostics, and Software Reliability (DIDSR), Office of Science and Engineering Laboratories (OSEL), Center for Devices and Radiological Health (CDRH), U.S. Food and Drug Administration (FDA)
Berkman Sahiner, PhD, Lab Leader, Division of Imaging, Diagnostics, and Software Reliability (DIDSR), Office of Science and Engineering Laboratories (OSEL), Center for Devices and Radiological Health (CDRH), U.S. Food and Drug Administration (FDA)
In sharing details of this important session, SIIM CMIMI planners offered these objectives:
1) Summarize recent initiatives for the regulatory evaluation of AI/ML-enabled medical devices; 2) Describe some of the current research projects from the AI/ML regulatory science research program at OSEL relevant to the design and evaluation of AI/ML-enabled medical imaging devices; and 3) Discuss regulatory science issues in imaging raised by natural language processing methods and large language models.
A valuable summary of the program, provided by event planners, follows: As more sophisticated techniques are introduced into artificial intelligence/machine learning (AI/ML) in medical imaging and as new types of AI/ML-enabled devices are tested for the market, new regulatory questions arise. The AI/ML regulatory science research program at the Office of Science and Engineering Laboratories (OSEL) at the FDA performs regulatory research to fill gaps in regulatory science by developing robust test methods and evaluating test methodologies for assessing AI/ML performance both in premarket and real-world settings to reasonably ensure the safety and effectiveness of novel algorithms. In the first part of this presentation, the presenters will review areas of emphasis: regulatory science tools and current research projects from the AI/ML regulatory science research program. In the second portion, they will focus on the regulatory science issues in imaging raised by natural language processing (NLP) methods and large language models (LLMs). Lastly, presenters will outline potential areas of application of NLP/LLM in FDA-regulated devices, and seek input from the audience about what types of tools would facilitate the development and evaluation of such devices.
Additionally, SIIM is applying for up to 14 hours of CE credits for in-person participation only.
For more information: www.siim.org