I’m excited to speak at the #ACR2022#RFS on the importance of non-interpretive #AI.
I hope you’ll join me live. If you can’t be here in person, follow along for the concepts in this 🧵
How do we get from today’s era of stupid computers to our future where one #AI Doctor can care for an entire ship of people #ACR2022#RFS
To get from here to @StarTrek, it takes lots of annotations, lots of coding, and lots of testing #ACR2022#rfs
Unfortunately, while we are waiting for all the annotating, coding, and testing, radiologists are frustrated and getting burned out #ACR2022#rfs
What are our most pressing needs today. HINT, for the most part it’s not interpretive tasks. Radiologists are really good at those #ACR2022#rfs
We need to simplify tasks for all people affected by the imaging value chain #ACR2022#rfs
Non-interpretive #AI is an algorithm where the primary output is not a diagnosis. Let’s explore some examples #ACR2022#rfs
Problem 1: How do we get rid of all of these reminders in our reading room? #ACR2022#rfs
Ideally, our dictation system would provide recommendations for structured language and staging systems #ACR2022#rfs
Problem 2: How do we balance radiologist schedules so that we can adequately take care of our patients and manage the clinical case load? #ACR2022#RFS
If you change one radiologist on the schedule, your day can go from smooth sailing to… #ACR2022#RFS
… a disaster with phones ringing off the hook that makes everyone want to scream or slap themselves silly. #ACR2022#rfs
Problem 3: How do we avoid these crowded waiting rooms? #ACR2022#rfs
Our #AI algorithms should be able to help us know when we are getting near the danger zone and help to notify patients so that they don’t sit in our waiting room #ACR2022#rfs
So, why aren’t we focusing on these non-interpretive tasks? #ACR2022#rfs