What Real AI Adoption Looks Like Inside Two Health Systems – Healthcare AI

In a latest webinar, Avi Sharma, MD, CIIP,  Jefferson Well being, and Leonardo Kayat Bittencourt, MD, PhD, College Hospitals, shared what AI implementation truly appears like in apply, together with the selections and classes that don’t normally make it into case research or information releases.

Moderated by Tom Valent, Aidoc’s Chief Enterprise Officer, the dialog explored the realities of deploying AI throughout complicated well being methods — from cross-specialty workflows and user-level adoption to governance fashions that may hold tempo with speedy change.

What Stood Out In the course of the Dialogue

1. The strongest sign that AI adoption is working? When clinicians discover it’s lacking.
Dr. Bittencourt shared an anecdote a few temporary system outage throughout which a number of radiologists flagged the absence of AI assist — not as a result of it failed however as a result of that they had come to count on it. The spike in nonengagement alerts throughout that window underscored how deeply the software had embedded itself into scientific apply.

2. Radiology should lead — or danger being sidelined.
Each audio system emphasised radiology’s essential position in enterprise AI technique. With distributors more and more advertising on to downstream groups, Drs. Sharma and Bittencourt argued that radiologists must leverage their deep information and expertise with AI to proactively assist form implementation, governance and cross-specialty coordination.

3. There’s no such factor as a one-size-fits-all workflow.
At College Hospitals, AI was built-in in a approach that allowed for user-level customization. Some radiologists most popular alert-based widgets; others used PACS-driven prioritization. The power to assist each was a key think about widespread adoption and an argument for methods to discover a platform that provides flexibility, not simply options.

4. Standardization has limits — particularly throughout hospitals.
Dr. Sharma famous that even inside a single system, deployment should account for native workflows, staffing fashions and knowledge environments. The shift from pilot to platform, he mentioned, required greater than technical integration — it demanded orchestration, scientific alignment and governance that would scale.

5. Governance remains to be evolving, and that’s OK.
Each Jefferson Well being and College Hospitals are nonetheless refining their governance constructions. Dr. Sharma shared how Jefferson Well being’s AI committee expanded from a radiology-specific physique to a multi-hospital enterprise steering group. Dr. Bittencourt described a tiered construction that features analysis, scientific operations and govt oversight with radiology on the desk all through. If you happen to’re constructing a governance framework, coordinating cross-specialty AI workflows or planning to scale AI past radiology, the experiences shared on this webinar might spark inspiration. Entry the recording.