“Those who can imagine anything, can create the impossible.” Alan Turing
This year’s Medical Intelligence Society (MIS) summer summit opened with a special tribute to all the clinicians on the frontlines who have sacrificed, and continue to do so, during this COVID-19 pandemic. The World Health Organization estimated that over 115,000 healthcare workers have lost their lives around the world since the beginning of the pandemic.
The first day’s activities started with an update and a primer on artificial intelligence in clinical medicine and healthcare. Hot topics included: federated and swarm learning, virtual twins, direct to consumer healthcare apps, extended reality, drug design and repurposing, digital biomarkers, and cognitive architectures.
There were many outstanding talks during the two-day gathering, but a few highlights of the meeting included:
– Dr. Fatme Charafeddine, Lebanon’s first pediatric electrophysiologist, who applied artificial intelligence in her homeland where revolution, pandemic, and other unrests have not deterred her efforts to learn AI and to deploy this resource for automated interpretation of pediatric EKGs.
– Jennifer Cortes and Kenny Leung of the Carle College of Medicine with their engineering-based medical school capstone project on improving epilepsy monitoring with AI-powered wearable device with a CNN/RNN hybrid AI strategy.
– Ji Lin of MIT speaking on the next step in the evolution of cloud to mobile AI to “TinyAI” (in the form of a neural net embedded on microcontrollers, or MCUs) and efficient deep learning for future healthcare wearable devices.
– Tim McLerran (presenting on behalf of the MIS graph database working group) and his brilliant presentation on the concept of using graph databases in healthcare to maximize knowledge extraction in healthcare data.
– Hatim Abdulhussein of NHS and Health Education England and his talk on Digital, AI, and Robotics Technologies Education (DART-Ed) that provides the NHS workforce with the knowledge in these emerging areas of healthcare.
– Ripon Chakrabortty of University of New South Wales in Australia discussing the impact of swarm intelligence optimization techniques for IoMT and IoHT with data grouped using automatic clustering procedure called artificial bee colony (ABC) optimization.
In addition to the short talks, there were two open forums: the first one was on some of the nuances of AI education: what is more important for clinicians is to learn how to adopt AI for clinical medicine and less important is learning to program.
The second open forum was focused on forming an AI center of excellence with various panelists detailing their institutions’ journeys. The meeting concluded with two AI workshops: one on use of convolutional neural networks and the other on natural language processing.
My personal observation is that AI in clinical medicine and healthcare is exponentially increasing in sophistication and deployment in some centers. The work by this group of clinicians in AI in medicine is especially inspiring during this very difficult time of our clinical careers.