Chairside Clarity: Using AI to Drive Patient Understanding and Case Acceptance

Gordon Barfield: Good evening, everyone. Gordon Barfield here from Overjet, and thanks for joining Dr. Prentice and I this evening, after what undoubtedly has been a long clinical day for many of you. So thanks again for joining, and I think we've got an exciting program for you tonight. We're going to bring you some real-world dentistry in almost, or near real time. We'll call this chair-side clarity, utilizing artificial intelligence to… Drive patient understanding and case acceptance in your practices, and we really have one of the experts on that in Dr. Prentice, and we'll introduce him to you in a moment. For those of you who were registered initially for this program on September 23rd, thanks for coming back. I apologize that we did have to move it to tonight, October 2nd, but the program will be the same, and we'll have all the same energy into it and tidbits for For you folks on the call, so… without further ado, my good friend, Josh Prentice. Josh, thanks for being on tonight again. It's always great to talk to you. We always have wonderful conversations, and I want to get this audience to know you a little bit. Tell me a little bit about you and your practices, and what kind of things charge you up about dentistry.

Joshua Prentice: Thanks, Gordon. No, great to be here. Thank you for the invitation. My name is Josh Prentice. I am a dentist in Flower Mound, Texas. I graduated from the University of Oklahoma about 12 years ago, and one of the things that got me interested in dentistry was Technology and how it's changing healthcare. So after working as an associate for 3 years, I bought a practice in Bend, Oregon. It was beautiful, gorgeous, amazing, and wonderful, had a great time there, but now that we have 4 children, my wife and I decided to move back closer to home, and so that's what led us to come to Texas about 2 years ago. So, what keeps me fired up and fueled up about dentistry is, you know, I'm… I love technology, I love… utilizing it to advance healthcare. Even if it wasn't dentistry, I'd be finding some piece of technology to advance some other sector, so that's what keeps me engaged. And it's a constant, you know, evolving and changing scene in the dental space, and it keeps me engaged and looking forward to coming to work every day.

Gordon Barfield: Love that. I was very similar to you, Josh, when I was practicing, and you know, the concept of change over time and not being stagnant really does keep us getting up in the morning, and I think that that ties right into where we are in dentistry today, right? This is an amazing time to be a dentist for so, so many reasons, not the least of which are artificial intelligence. So, let's kind of jump into that today, and let's give first the audience a little bit of a roadmap, the things that we'd like to cover today. Again, talk about clinical clarity, right? For making the diagnosis. for us as clinicians, clarity comes right to the forefront. And then I think also about understanding from our patient. They need to have clarity, and I think that AI helps us do that in a very, very unique way. We'll talk about a few tips for communicating artificial intelligence findings. and a diagnosis to the patients, and Dr. Prentice will give us some of his best use cases there around communicating this stuff. And then a little bit about the business impact. We'll ask Josh how AI has changed his practices over time. So, what do you think, Josh? Let's dive right in.

Joshua Prentice: Sounds good.

Gordon Barfield: So, this is one that I got from my friend Keith Baines up in New Jersey. This is a clinical series that he sent me recently, both a clinical set of photos, as well as some x-rays that go along with it. You know, the thing that jumps out at me here, Josh, and we all kind of suffer this fate in practice, is making the diagnosis when there are a lot of things. that can affect the diagnosis. And so, when you look at something like this, a routine bite wing, you know, it's obviously an adult patient, they've had some work. What are a few of the things that you think about when you come up against this wall of, how do I actually make an accurate diagnosis on something like this?

Joshua Prentice: Yeah, you know, and it's… I look at this x-ray, and I go. when am I looking at this x-ray? Is it at 8 in the morning when I'm fresh in charge, or is it Thursday afternoon at 4.30 when I'm ready to head out the door? You know, is it with Ms. Jones, who you've seen for 10 years and you have an emotional connection with, or is it a new patient? You know, with clarity and diagnosis, I think we have to be honest, and a little bit of introspection is warranted, where we have to admit, we're going off our own training, our own knowledge and experiences, and so for me, I look at this x-ray, and I go, oh, I wish it had Overjet. For me, you know, I say that kind of half-joking, but half serious, you know, I said I had a practice in Oregon, and I used Overjet for 5 years, and absolutely loved it, and when I came to touch I switched practice management software, and there was a 6-month period where I didn't have, any overjet tied into my x-rays, and you know, I think something that a lot of people don't talk about that flies under the radar is just the clinical fatigue in decision-making, or the fatigue that you have in decision-making all day. So I look at this x-ray. you know, sitting here Thursday afternoon at 4.30, or 5 o'clock now, and I go, I am so thankful for the clarity that Overjet provides. And I know it's a bit of a plug here, but it's sincere, so…

Gordon Barfield: Yeah, and you know, it's interesting, you look at something like this, which is very, very typical, right? But it has some features in it that really tend to get us off track, not the least of which are the previous restorations. You know, those things are sitting on that X-ray, these radio opacities. and they're shining like a BMW headlight in our eyes, and it's well known that when we have existing restorations. That dulls our visual acuity. in looking at very subtle grayscale changes, right? So, in this particular instance, Keith was sending me this, and he wanted to note something on tooth number 20 here, and so for the audience, if you take a look at that, you know, if you're like me, I looked at this thing, and I said, I'm not really sure if there's any disease on this x-ray or not. It looks fairly normal, it looks like it has some good dentistry done, for sure, and would I just skip right by this on a quick exam where I had you know, four hygienists waiting for me, and I was busy, and maybe it was Thursday afternoon, like Dr. Prentice said, but… to his point, the utilization of artificial intelligence to kind of highlight something like this can be very, very helpful, because here, essentially, this is drawing our eye in an area to look, and I like to think about it, you know, Josh, we've talked about before. this disease probability map, Overjet not telling you what the condition is, or that there is a cavity here, but that if you look in this particular area, you might be more likely to find one than another area. What are your thoughts on that?

Joshua Prentice: Yeah, absolutely. I mean, you can't blindly trust that you still need to use that clinical decision-making, to, you know, to make the diagnosis, but it definitely helps point you in a direction and maybe catch something that you might have missed.

Gordon Barfield: Yep, for sure, for sure. And so, Keith was gracious enough to send me the clinicals on this, and as you guys in the audience can see, there's clearly dental carries here, probably a D2 into the distal surface of number 20. So, you know, he says that AI helped him at least Raise his suspicion level about this, and then look at it a little more carefully, and it looks like he had a good outcome here for the patient. Now, you know, Josh, this one's a little different, isn't it? Most of the audience is going to be able to see dental caries here, but there are some other questions that they would need to answer around this particular exposure, this bite-wing x-ray. What are the things that you think about here?

Joshua Prentice: You know, looking at this, I go, well, you know, there's some lesions that I'm suspicious of, that I would suspect have cavities. There's also some overlap in some of these teeth, which makes it more challenging to read. You know, there's definitely a few areas and contacts that don't have really, really clear imaging, so it's nice to take a second shot here, and then If you do have a second image, and you do have the benefit of having something like Overjet, or whether you do or not, having that second image is really going to be helpful in these interproximal areas that may be a little bit hard to read otherwise.

Gordon Barfield: Sure. And so, just from a visual perspective, you know, kind of this is the AI output. And I'm curious, you know, we were talking a moment ago about the concept of clarity, and clarity as it ties into communication. you know, it's one thing for us to be able to have a map here to say, hey, look here preferentially, you're more likely to find disease. How do you go about actually transmitting that idea to a patient and Do we even need to, or are we in a different era now in terms of communicating with patients, now that we have quantifications and colors on x-rays? What are your thoughts there?

Joshua Prentice: Yeah, you know, I think… and we'll get to the business aspect of it later, but I have seen in two separate practices, in two separate states, that my case acceptance goes up when I have overjet, and I just truly believe that patients feel more engaged. I like to call it co-diagnosis, so they can understand what you're looking at, what you're seeing, and make a decision together, and it brings some ownership and some insight, to an area where patients a lot of times have just been looking at shades of gray and they don't really know. You know, so I think that in terms of, clarity with, you know, it also, provides clarity for your team. whether you have associates, you know, you've got four hygienists, or you've got two, it also kind of helps standardize the team, so we're all kind of focused on the same thing. We all have a little bit of quantification and some different annotations that help us differentiate between different lesions, and ultimately, you know, circling back, it gives clarity, which then, I believe, ultimately ends in more case acceptance and patients receiving treatment that otherwise may or may not have moved forward, based on someone's own individual communication skills.

Gordon Barfield: Sure, and that's the key, isn't it? That's really what the goal is, is to get the treatment that the patient needs. It's interesting, you know, we see a couple of different lesions here on 28 and 29, And they look different on the Overjet overlay, and we're going to look at them clinically in a minute. What are the things that you think about from a, you know, from the clinician's eyes when you see that image, on 28 and 29, where the two approximating lesions are?

Joshua Prentice: When I look at it, I go, it would be, like I mentioned earlier, it'd be great to have a second image, but if you didn't have a second image, you know, this provides some insight, and in my experience, I know I'm biased, and I've been using the software for 5 years, but I definitely feel like it helps with you know, diagnosing and also kind of drawing your attention in that area so that you can take a deeper look. There's also a lot of tools, whether it's with your native imaging software, but also Overjet does a good job of enhancing images and making them easier to be read, so that's another kind of component of Overjet, which often goes missed, in my opinion, so… Just using those tools to enhance this image to get the best, best diagnosis you can.

Gordon Barfield: Let's take a look at it clinically here. For the audience. there you can see Dr. Baines has opened up this… these lesions, and you can see they're into the dentin, of course. And, you know, that brings up, Josh, an interesting point that we were talking about just before we jumped on today, which is that… that concept that we all were trained on in radiology But we oftentimes forget, right, about the depth of dental carries compared to what we can see on a dental x-ray. And that's, that's what, what is it, about 30% more? You know?

Joshua Prentice: That's… I remember that question back from dental school, and yeah, it's, you know, I can't remember… I can't pull out the study, but yes, whatever you… I kind of tell myself, remember I'm from Oklahoma, if the… if the… if you can see it on an x-ray, it has to be bad enough. But we know that, clinically, the carries are expected to be 30% beyond what we see from an x-ray.

Gordon Barfield: Got it, and I think we're going to see a little bit of that in a moment when we tie into some of the real-time dentistry. So, without further ado, let's kind of jump in for the audience, and this is… Dr. Prentice was gracious enough to bring along a case that he did this morning. Am I right about that, Josh? This morning?

Joshua Prentice: That is correct. This was my 8 AM patient this morning.

Gordon Barfield: And so walk us through what we're looking at here, right? We took some bite wings, I guess it was a checkup examination, or what have you, and what are the things that we think about when we see something like this, 8 a.m. in the morning, and we're going to actually take this step by step, and we're going to look at this through to conclusion.

Joshua Prentice: Yeah, so again, this was a patient that came in… he came in through hygiene, and we were taking our routine series of x-rays. This was one bite wing that we had over on the left side, and you know, to deepen the story, I've got a, I've got my four children, my son plays on this father's, you know, baseball team, and we've had, you know, all kinds of experiences, both at the school and with sports, and so we're having this conversation here in the practice, and, you know, I do think there's there's the emotional side of this, where you're like, oh, I don't know, this is my friend, like, what, you know, what would we do? Or, you know, how do we read these x-rays? Not to say that emotions get… into your diagnosis, but they can influence your decision making. And so, I looked at this, and I said, hey, there's a couple areas I'm suspicious of. So, this is the initial x-ray. You know, I look and I say, yeah, there's a history of some cavities in here, there's, some previous work, there's a few areas that I, you know, I kind of have my eye on. But, you know, I kind of know how this story ends, and there's definitely some spots that I Could have missed without, you know, my eyes being pointed in the right direction, but there's also spots that the software may not pick up on that, as a clinician, you gotta pay attention to and make sure that you're not blindly trusting the software.

Gordon Barfield: So let's take a look at that AI read, right?

Joshua Prentice: Yep, let's go So this is the annotation from Overjet, and from this treatment plan, we decided to move forward with the restorations, or with fillings on the incipient lesions, according to the software here, because I also know that, you know, I expect these lesions to be 30% beyond what I'm seeing clinically, or what I'm seeing on the x-rays here. So for me. The deciding factor, a lot of times, is if Overjet flags a lesion as an incipient lesion here in yellow, and it's 50% through the enamel, which is easily highlighted in white. I recommend moving forward with treatment because I know we're going to be closer to that dentin than is portrayed on this x-ray. All of these lesions were more than 50% through the enamel, so we decided to move forward, and we have the benefit here of the next couple slides seeing how that played out. So, we took a lot of pictures this morning, because I was like, this is a great, a great example case for this afternoon. So, Phil, Gordon, you can click through those. This is the canine. This is… these are all pre-op photos, so this is, number 1213. This is tooth number 18 here. And this is between 20 and 19. So, we're gonna start right there between 20 and 19, so looking at our annotation over here on the left, again, incipient lesion detected on the distal of 20, but as you can see on 19, even though it says it's an incipient lesion that's still in the enamel on the x-ray, we can tell that that has gone into the dentin, so that, again, has progressed more than was portrayed on Overjet's imaging there. And that's not a dig at Overjet, it's just something that we need to be mindful of based on our training. So, moving forward. Cleaning out that cavity. Got all the way down to the bottom there, got clean margins, everything looks great, went ahead and filled that area. So then we're gonna move to the next area, which is between 19 and 18. So, we had recurrent decay under a previous restoration on the mesial of 18, and then it looks like a larger a larger lesion on the distal of 19 than was annotated on the mesial of it, and I look at that and I go, it's going to be 30% beyond that, and we know once it gets to the dentin, it can go a lot faster. So, as we progress through this series here, we go, yeah, absolutely. It was further along. And we detected initially, or then was annotated on the x-rays, and that's where it ended up clinically. So, a reminder that, you know, had you just looked at this and said, this is, you know, it's not into the Denton yet, we're gonna watch it. This is what you would have been watching on a friend who may not have been getting x-rays for another year until their next hygiene visit.

Gordon Barfield: And there's the recurrent decay underneath the restoration there on the mesial of 18. Josh, let me ask… let me pause for just a second here, and I want to ask, because this always came to my mind when I was in practice communicating this to a patient. This is a number of lesions, right? And so the treatment plan that we're going to deliver to that patient is oftentimes received as a significant treatment plan. You know, if you're not getting a cleaning and one filling, well, that's a lot of dentistry, and are they over-treating me? How do you, in this day and age, with AI, And with color and quantification. Leverage it to communicate that in a way that the patient is, let's call it more accepting of the assessment and the diagnosis.

Joshua Prentice: Absolutely. I mean, for me, I say this is an FDA-approved software that aids in the diagnosis of finding caries and periodontal disease. It is accurate, you know, based on the studies I've read, almost 94% of the time, so we still have to use it with a grain of salt. But it helps us detect lesions earlier and sooner than we have been able to in the past, and therefore, yes, we do catch areas before they become problematic. Once we move forward with treatment, I always take photos just like that. or just like this, and had I not been sharing these slides, or these photos with the presentation this evening, I still would have taken the same photos. So, having your patient… if a patient has, whether they have any hesitation or not, sharing these photos, if they're open to seeing them afterwards, really builds a lot of trust, and then they are kind of your evangelist, and they help refer friends and family, because they… Appreciate you're using the latest technology that's available, and you're catching things early, and you're engaging them in their treatment plan, whether it's the Clarity and communication at the diagnosis, or whether that's involved during, you know, during the actual procedure.

Gordon Barfield: Love it, love it. Involving the patient more, and now more than ever, I think it's easier to do that, isn't it?

Joshua Prentice: Yes, absolutely, and I will admit, like, we don't always throw it out there that, like, hey, this is our AI. We say, hey, we've got some new imaging software that helps us catch things early, because to some patients, you have to… you do have to judge that. Some patients, you know, AI may be the big boogeyman, but a lot of people are becoming more and more familiar with it, and seeing the practical applications of it, and becoming accepting of it. But, you know, just communicate with your patient the way you'd like to be communicated. communicated with.

Gordon Barfield: Yep, love that. And so, I guess this is the clean-out here.

Joshua Prentice: Yep, clean out there, and then filling. Ta-da!

Gordon Barfield: Great work.

Joshua Prentice: Alright, thank you. This, we're moving up, this is the mesial number 12. kind of same process that y'all all know how to do, so, cleaned out. This is an interesting image. So, if you notice, there is no distal lesion on, on number 13. However, I know a few slides back, we had the original image, and I said, I believe, or sorry, I misspoke. There's no lesion on the distal of 12 here, annotated on Overjet, but clinically, I made the decision that I believe that there's a cavity on the distal of number 12, and we went ahead and treatment planned it, and I told my friend, I said, whenever we go in and we treat the mesial of 13, I'm going to look visually at the distal of 12, and this is my camera looking there, and I said, yeah, sure enough, there is a spot on the Dyssel 12 that I was suspicious of, and so we went ahead and treated it here. So that's what it looked like. And then, when it was cleaned out. And… build. So, just a reminder that you can't blindly trust AI. It's not going to do your clinical decision-making. You still have to use, you know, use your training and education.

Gordon Barfield: That's a really good call-out. One of the things that we often talk about is we don't want to ever fall into the habit of looking at a color on an x-ray and equating that to the need for a filling, right?

Joshua Prentice: Absolutely. Following up, this is the distal in number 11, just basically how it appears, you know, again, you look at that and it goes, it's almost to the dent in there, and then clinically you find out it isn't to the dentin. So, just… this was just a good case, and I asked my buddy, I said, hey, would you mind if I took all these pictures and used them today? He said, absolutely. So, cleaned it out.

Gordon Barfield: really does highlight the nature of these… what we think are early lesions, but ultimately turn out to be lesions into the dentin, right? And, as you know, when they get into the dent, they start spreading and they go faster.

Joshua Prentice: Absolutely. Going there.

Gordon Barfield: Really great work, Josh, that's amazing. I wish I could have done composites like that when I were in practice.

Joshua Prentice: Thanks. I'd say I'm always up for room for improvement, so I throw this out there, and I'm humbled by some things and everything else, but at the end of the day, I think this is a nice side-by-side x-ray. We took an x-ray afterwards to show, like, hey, we did catch these lesions early. Y'all came along on the journey here to see, you know, what they looked like clinically, and I'm happy that I was able to provide this service for my friend. he was able to see the technology, you know, and then engage him during the treatment, so I appreciate him letting me share this with y'all. Amazing. And then we took one more x-ray just to, you know, go back a little more between 18 and 19 there, and it even caught that extra piece of composite I left there on the… on the distal of number 19, so I'll be sure to have my hygienist go back there with the scalar.

Gordon Barfield: Got it. Amazing the things you see, right?

Joshua Prentice: Yep, you can't unsee it once you see it, so, yeah.

Gordon Barfield: Exactly. Exactly. Before we jump into a little bit of Q&A, we'd love to give the audience an opportunity to ask questions of you, Dr. Prentice. You can share, you know, this expertise that you've obviously got, and I've grown to really appreciate since I've known you. Talk a little bit about the effect of AI, specifically Overjet, on your two practices from a business standpoint. can.

Joshua Prentice: Yeah, absolutely. So, you know, I have the benefit of having a nearly fee-for-service practice in Oregon, and then a PPO practice in Texas, so I've seen it work in both different worlds, or in different states, but I will say that adding Overjet to each of my two practices Grew our collection and productions by 30%. And I would say that's not… a lot of that is related to that communication with patients, being able to clearly communicate with patients, meet their expectations, co-diagnose with them, and involve them in the treatment. We just… we have more case acceptance, and so, yes, that has grown two practices in two states by 30%. Amazing. I mean, at least. I know, I measured it with 30% in Oregon, and I was there for 6 years, and here in Texas, I've been here for 2 years, and it's contributed at least that amount.

Gordon Barfield: you know, 30%'s a lot. You could work the same amount and change your lifestyle pretty dramatically on 30%, or the other alternative would be, you could work a little bit less and make the same amount of money, if you chose to do that, right? So you could look at it different ways.

Joshua Prentice: Yeah, and I'll say that one of the biggest things, and like I said, when I got here in Texas, there was a 6-month period I didn't have this, the software, and it's… obviously, I can practice without it, but I think, Gordon, I think you mentioned it, somewhere in the FDA filings for this, which is a huge… thing to have FDA approval for, all the clinicians found more cavities with the software than without it, and it's just… it's… it's a second set of eyes, it draws you in the right direction, and we just catch things earlier. I think that is huge. It aids us, it helps standardize treatment planning, it builds a, you know, a community that's going to be more accepting and trusting of our findings and our diagnoses. And at the end of the day, too, at the end of a long week, there's less… just mental fatigue from all the nuanced decision-making of looking at black and white x-rays. So that's… again, it's one of the unsung heroes, but I absolutely love the software, love what it's done for me, what it's done for my practices, and I… obviously, that's why I'd love to… I'm more than happy to be an advocate for OpraJet.

Gordon Barfield: That's amazing, Josh, and also, congratulations here. This is such good quality dentistry. I love highlighting you know, great dentistry, and we can never get enough of that, right? So, beautiful work here on the clinical stuff, and I certainly appreciate you coming on and sharing these insights with the audience. This is a conversation, I think, that's very, very pertinent right now in dentistry, and will grow more pertinent as we move along toward more involvement of AI in every part of dental practice. This is just the beginning. We're kind of at the starting line here. So, let's kind of open this up a little bit for questions. Dr. Prentice and I would love to take any questions that you folks in the audience have. You certainly can post those in the chat. We have a monitor who's monitoring the chat, and Olivia can pass those along to us if you have any.

Olivia Prater: Yeah, there's two that's come through so far. We'll start with the first one I see here. It says, if AI makes a mistake, can we adjust the segment shape?

Gordon Barfield: Go ahead, Josh, you've got that one, and if you don't have that one, I'll take it.

Joshua Prentice: I know early on, and I have switched imaging software, but I know early on, Overjet did take in feedback, and you could turn findings on and off and everything else, but I'm gonna let you speak to that one.

Gordon Barfield: Yeah, you're right about that. So, to the specifics of the question, yes, the clinician, if they don't agree with any AI finding, that can be deleted from the data set, taken off the x-rays, and you can provide that data back to us at Overjet, and in that way, be involved in training and developing the software and the technology to be better and more precise over time. And if you do choose to provide feedback in that way, it is very, very helpful to us. Our clinical team validates it. our machine learning team validates it, and then we bucket it for use in retraining exercises down the road. So, in a small way, you'd be helping organize dentistry Mature this technology for the betterment of patient care. And to the second point about the segmentation, yes, you can change the outline of the cavity segmentations if you want to do that. It's a little tricky to do. We can show you how to do that, but it is possible, and you can change the outlines of your periodontal bone level measurements, so forth and so on. So, the most important take-home around that is that the clinician is in full control of not only the assessment and the output of the AI, but also what they show the patients, right? So, you can control the narrative there. There's never a time where the AI is going to override anything that you determine is a disease entity. You're fully in control of that. So, we feel like at Overjet, human-in-the-loop performance is critical for this. I think that will always be the case. It is a Class 2 medical device, as most of you know, and there are regulations around that, so hopefully that was helpful. And I think, Olivia, you said maybe we had a second question?

Olivia Prater: Yes, second question here, it says, does Overjet detect features on intraoral, too?

Gordon Barfield: intraoral photographs, I would expect that the caller is asking about intraoral photographs. That's an interesting question, because computer vision in its current state could be used to read a lot of different things, right? Including pictures, right? Google reads our face off the internet by looking at pictures of it. And so the answer is, current state, no. However, you might not be surprised if I told you that we are working on those things, currently at Overjet, and up to and including things like multi-image synthesis. So, again, the technology is at the starting line, I think, and it will only get more sophisticated, better, and more useful as time goes on. Thanks for that question.

Olivia Prater: And that's all we got. Nothing else has come through. I know we're at time. If anybody has a last-minute question, go ahead and type it in, but, thank you for everybody who submitted.

Gordon Barfield: And I would say, while we're waiting on that, Olivia, thank you, Dr. Prentice, so much. This is always fun. It's always a highlight of my day to talk to you about clinical dentistry, and I think I speak for the audience in saying that I would be surprised if any of them did not get a lot out of what we just did. You know, this is really real-world dentistry, isn't it? It's the application of technology to improve patient outcomes, and you were gracious enough to bring along a case that you did this morning. So, that's amazing. I mean, you saw a full book of patients, and then put an entire slide deck together. I don't know how you do it.

Joshua Prentice: Well, I appreciate the invitation, and I'm happy to share my passion, so thanks for having me, Gordon. Thank you for Olivia, for being here.

Gordon Barfield: Thanks so much, and for you guys in the audience, thank you for being here, and hope everyone has a wonderful evening.