Finding His Own Rhythm With Rad AI Reporting

At Radiologic Associates of Fredericksburg, Salvador Trinidad, MD, describes how personalized coaching helped him refine his reporting routine.

When Salvador Trinidad, MD, began using more of Rad AI Reporting’s advanced AI capabilities at Radiologic Associates of Fredericksburg (RAF), the change initially created some inefficiencies.

“I needed Rad AI to meet me where I was,” he said. “I needed to find out what it could and couldn’t do and then make some adjustments to bridge any gaps. I had to learn how to use words that would allow the AI to drop my words in the report at the appropriate place. It also had to learn from me.”

Alongside those dictation adjustments, Dr. Trinidad worked on making the controls easier to use.  

“Reducing the clicks was also important, so I programmed some buttons and updated my macros. Initially, incorporating the AI slowed me down, but it made me more thorough. As I became accustomed to my new workflow, I increased my efficiency, eventually surpassing my old workflow in speed and accuracy.” 

Personalized coaching supported that process. Through Rad AI’s Impact Accelerator Program, Dr. Trinidad and several of his RAF colleagues participated in personalized coaching sessions with Maribeth Perik, a lead implementation product specialist. By observing the workflow in real time, Perik could troubleshoot problems as they occurred and tailor the sessions to Dr. Trinidad’s priorities.

“As we interacted, I really kept bringing it back to what my pain points were,” Dr. Trinidad said.

Perik described her approach simply: “We have to meet you where you are and then show you some other opportunities. And if they stick for you, great. If they don’t, we find something else.”

Changes He Could See and Measure

CT accounted for the largest share of Dr. Trinidad’s workload, making it a natural place to establish a routine before expanding to other exam types.

“I wasn’t trying to make too many changes at once,” he said.

As the routine became more familiar, the gains became apparent. Manual template selection fell from approximately 28% to 7%. Use of AI features grew from essentially none in August 2026 to approximately 20% of all reports during early October, including 36% of CT reports.

An internal analysis found that Dr. Trinidad’s measured reading time decreased by 14% between September and early October. The result reflects encouraging progress as he refines his workflow, even as he continues to work across two reporting environments during RAF’s transition.

Improvement extended to examinations completed without AI features. There was even improvement in reading time for a chest X-ray since starting the program. These gains highlight why supporting adoption means looking beyond new technology to the habits, templates and manual steps that shape the whole reporting workflow.

While the metrics captured the efficiency gains, Dr. Trinidad noted that the numbers didn’t tell the whole story. As he put it, “The other thing that the stats don’t show is the report quality; I was able to implement it in a way that gave me a better quality report.”

Dr. Trinidad’s experience adds an individual perspective to RAF’s earlier case study, which documented throughput improvements during adoption and a positive association between greater ambient reporting use and radiologist throughput.

Less Rechecking, Less Stress

An abnormal or interesting finding could interrupt Dr. Trinidad’s usual reading pattern. Picking up where he left off sometimes meant checking a section again to confirm it had already been completed. Visual cues in Rad AI Reporting helped keep track.

“As soon as you say something that goes under that heading, the text changes color,” he said. “If something stayed in the original color, I knew right away, ‘I haven't mentioned that structure yet, let me double check.’”

The cues also made it easier to recognize what had already been reviewed:“When I use the AI, I was able to quickly see, ‘I touched on everything, so I don’t need to go back and look at things again.’”

That difference affected how the work felt, too.

“On the high-volume scans is where I found it most beneficial, I certainly was less stressed,” Dr. Trinidad said.

RAF President Roni Talukdar, MD, has described his own experience in similar terms: “I see a decrease in my mental load. I see an improvement in my day to day. I see myself being more efficient.”

What Comes After Go-Live

At RAF, which has grown to more than 70 radiologists and reads for 40 sites across four states, the transition to Rad AI Reporting is still underway. In November, Dr. Trinidad reported splitting reading time approximately equally between Rad AI Reporting and PowerScribe® while covering different facilities.

After 18 years of using a different dictation software at one facility, the system and its features were well known. Even so, switching between platforms made the absence of Rad AI features noticeable.

“I miss the ease and efficiency of the new system,” he said.   

The coaching continues, as Dr. Trinidad refines his workflow on other case types, with Perik using performance data and his feedback to guide their next steps.

“It’s nice to have somebody who’s interested in making things better,” Dr. Trinidad said.

Learn more about Rad AI Reporting.