Discovering AI Together
ACR’s RFS delves into deep learning as a community through a virtual journal club.
Data science is revolutionizing the way medicine is practiced. Machine learning application is on the rise in just about every field of healthcare, signaling changes that have some specialists, including radiologists, speculating on how the ever-improving technology may change their position in the landscape.
The ACR RFS has been delving into the current state of data science in radiology through its AI journal club, started under the auspices of the newly-formed RFS AI Advisory Group. The club, which launched in December 2017, was the brainchild of ACR RFS Vice Chair Daniel Ortiz, MD, who wanted to create an online discussion forum to democratize AI in radiology and equip RFS members for success in the advancing IT landscape.
“A lot of emerging technologies are coming out, and most of the residents don’t really get exposed to some of the high-impact and cutting-edge techniques and tools,” says Ortiz, chief resident at Eastern Virginia Medical School, in Norfolk, Va. “I come from a pretty small program that doesn’t have a lot of high-end resources to be actively engaged in newer technologies like AI.”
According to Ortiz, the AI journal club’s January gathering was a record-breaker across all RFS journal clubs thus far. The session garnered 340 registrants, 145 attendees, and over 500 views of the video recording within seven days of being posted online. Judy W. Gichoya, MBChB, MS, chair of the RFS AI advisory group which organizes the AI journal club, explains the appeal: “We wanted to get an online community together that could help non-technical residents gain technical skills,” says Gichoya, an informatician and radiology resident at the Indiana University School of Medicine. “AI is a hot topic right now, and it’s going to affect how radiologists practice. It’s not being taught in school, so this club gives residents a good avenue to get involved.”
Both Gichoya and Ortiz agree that participating in any journal club affords residents the opportunity to be exposed to new topics not traditionally covered in coursework or day-to-day training. In the case of the AI journal club, residents see what new technologies are being developed and get beyond the hype and anxiety.
“Most radiologists are aware of AI given the amount of discussion about it at RSNA meetings and in the College, but most of them don’t have a real understanding of what it is and how it can be applied,” says Gichoya.
Journal clubs are traditionally siloed by specialty, but the AI journal club is looking to create a broad and diverse community of expertise including programmers and data science students, in addition to radiologists. According to Ortiz, on one side of the spectrum of attendees are the radiology experts who know the workflows and details of imaging and are looking to peek behind the curtain of programming. “They want to dabble in and get exposed to machine learning in the same way that most radiologists understand medical physics,” says Ortiz.
On the other side of the spectrum are the people who will be the future developers of these algorithms, the problem solvers and developers of use cases, notes Ortiz. “The radiologists in this cohort might not have exposure to the technology in their residency programs, but virtually connecting to experts in the field nationally might stir an interest in them to explore beyond the scope of their departments and programs,” he says.
Besides increasing knowledge about a specific content area and improving critical thinking and appraisal skills, at the heart of the RFS AI journal club is a sense of community-building. “In our view, a community that learns together, grows together,” says Gichoya.
By Maria Qadri, PhD, freelance writer, ACR Press