How AI is Making Radiology Better

Recorded live at the Imagine Client Conference in Las Vegas, Dianne Keen sits down with Michael Brozino, Co-Founder and Chief Commercial Officer at Maverick Medical AI. Brozino came up through radiology's last great transition, eleven years at Kodak as film gave way to filmless and eleven more at McKesson, and he uses that history to explain where text-based AI actually fits. Not the vision AI that reads the image, but a model that reads the unstructured report the way a coder does, pointed at the part of the business that is repetitive, high in volume, and expensive when it is wrong.

From there the conversation moves from throughput to workflow. Brozino describes practices running in excess of 85% direct to bill, and reports coming back with their codes in under two seconds. The part that matters most to operations, though, is the flag: catching missing documentation at the point of dictation, while the radiologist is still in the exam, rather than chasing an addendum days later when they have moved on. Keen and Brozino also cover scaling a coding operation through acquisition without scaling the team, where AI already sits across the front, middle and back of the revenue cycle, and why the guardrails are still being written.