"Doctor, Keep Your Eyes on Your Patient": One Radiologist's Unexpected Revelation About Rad AI Reporting

After 30 years in practice, Ken Ford, MD, a radiologist with American Radiology Associates, was accustomed to what radiology reporting felt like. For decades, he and his colleagues have been tied to their dictation software, dividing their attention between complex patient images and the constraints of traditional reporting workflows. 

Then he tried a different way of working with Rad AI Reporting. What surprised him most wasn't just speed or efficiency — it was something more personal.

"I realized I could keep my eyes on my patient."

In this conversation, Dr. Ford shares how changing his reporting platform drastically reduced physical and mental fatigue, and why returning to a legacy platform, like Microsoft PowerScribe 360, is no longer an option for him.

The conversation has been edited for length and clarity.

Why did you start using Rad AI Reporting?

Our group had already been using the Rad AI Impressions product, and when it became clear that PowerScribe 360 was sunsetting and would no longer be supported, we started looking for an alternative reporting platform. They asked me to be part of a team that tested the product and evaluated whether it could be a replacement product for PowerScribe 360. The verdict from everyone who tested it was an emphatic “yes.” 

What was the "light bulb moment" for you that it was a new way to report?

The real light bulb moment was when I realized that this product allows me to keep my eyes on my patient.

I likened it to the first time you connected your phone to your car and went hands-free. Before that, the phone was awkward, it covered your peripheral vision, you were fighting with the device. And then somebody said, "Connect it to your car." And you did, and you were like — I'm never going back. Now I can keep my eyes on the road.

The way we used to work in a traditional reporting platform, you have a templated report and you're constantly going back and forth between your images and the report, clicking into the next section, talking about it, clicking to the next, talking about it. In any single dictation, there are probably 50 changes in attention between the images and the report. Literally, your eyes are going back and forth that many times.

What Rad AI allows me to do is keep my eyes on the images. My eyes are always on the images. I'm speaking into the Omni Box, and at the end, I say "Omni Report," and it parses out all the different parts for me.

What does that mean for patient care?

In my opinion, the number one cause of medical errors in interpretive radiology is distraction. It breaks your concentration. It breaks your flow. You may forget what you were looking at or talking about. Things can fall through the cracks.

When you're doing it manually, you're clicking into each part of your template one at a time and talking about that section — and only that section. It segments your interpretation. Look at the liver. Stop. Report the liver. Look at the gallbladder. Stop. Click the gallbladder section, talk about the gallbladder. You're constantly interrupted.

Now I can just speak freely as I'm interpreting the case and switch freely between parts of the abdomen, parts of the pelvis. The system is just really good at parsing those out. It makes very few mistakes.

Don't you want your radiologist to be able to focus 100% of their attention on the patient? That's what we should want. And the problem is, radiologists don't know they can want that. That's the epiphany. That's the paradigm shift.

Let's talk about fatigue. That word comes up a lot in radiology right now.

By being able to concentrate on the images and not having to constantly look back and forth, in any single dictation I'm saving upward of 20 to 30 distractions. There’s nothing that contributes to stress for a radiologist more than distractions. 

I've found a definite reduction in stress and fatigue — mental, emotional and also physical. Neck strain. Eye strain. The thought of going back to that pattern, back and forth on every report all day long ... it's an unacceptable thought to me.

Legacy reporting systems are known to have lag. What has your experience been with Rad AI Reporting’s speed?

When legacy systems are integrated within a complex hospital environment, they often become slow and cumbersome — there's lag in voice recognition and toggling between fields. With Rad AI Reporting, there’s zero lag. I can dictate it as fast as I can speak.

With traditional PowerScribe platforms, every single field toggle means you have to count — anywhere from two to 10 seconds — before the cursor even moves. You can't get any rhythm going. You're constantly looking at the report thinking, "Is it going to take it? Is it going to move?" And then finally it does, and you speak the next field, and you wait again. Two to 10 seconds. Every single time.

That delay is especially brutal on toggle-intensive reports. Bone density DEXA studies, for example — there are so many fields. I used to pull my hair out all day, screaming at IT, "When are we going to fix the lag?" With Rad AI Reporting, I don't have that problem anymore.

How has the accuracy held up?

It's outstanding. The voice recognition is better than what I was used to. And it does a great job of parsing what you've said and placing it in the appropriate part of the templated report. I've found that it takes minimal editing — less than I experienced with PowerScribe 360 or PowerScribe One.

I can be reading a case, hopping between organ systems as I scroll the images. I can say something like, "Stones are seen in the gallbladder. Multiple liver cysts are seen. The uterus and adnexa appear unremarkable." I'm jumping all over the place, the way you naturally do when you're interpreting a complex case. And it will naturally parse those findings correctly into the right sections of the template. 

I've gained about 20% efficiency overall — I'm hitting my productivity numbers hours earlier than I was before. But beyond speed, I think I'm generating a better product, one that's not only faster and less fatiguing to produce but more thorough and more tailored to what doctors actually want. My impressions are more complete than what I was producing before. And what doctors want is an impression section that tells them what they need to know. That's what this produces.

What about radiologists who are hesitant to try something new?

I've been a practicing radiologist for 30 years. I've been through every one of these transitions. Every time, the hesitation was there. Every time, once you made the change, you were glad you did.

Humans shouldn't do what a computer can do better. This product can do certain things better and more seamlessly than we can as humans — specifically, taking spoken findings and organizing them into a complete, structured report. So, let it do that. Because it's more effective for us, and, at the end of the day, I think it produces a better product more efficiently.

What would you say to a radiologist who's on the fence about switching reporting platforms?

If you give Rad AI Reporting a try, you'll have less eye and neck fatigue, less emotional fatigue. And you will find it incredibly gratifying to be able to keep your eyes on your patient and your focus where it should be — rather than experiencing the innumerable distractions of going back and forth between the images in PowerScribe.

You get a little rush each time the report generates because you get to see the technology do real work for you, at a very high level. And once you've changed your paradigm, you don't want to go back to the old way.

Doctor, keep your eyes on your patient. Let the technology work for you.