99: AI in Dentistry: AI's Analysis of the ADA AI Whitepaper

Welcome, everybody, to another episode of the Dental Boardroom podcast. So, as you know, I am currently doing a series on AI, which is going to be about, I don't know, around 7 to 10, when all is said and done, episodes on AI in dentistry. And my first episode was an introduction. The episode before this one was an introduction on AI, generally speaking, and its application in. In dental offices. And a little bit of an intro into the companies that I'll be interviewing so that you can see who are some of the leading players out there in the space of AI technology in dental offices. So I hope you find it helpful. In the meantime, I want to show you one way that I'm going to use AI in order to display how effective and how powerful AI is. What I've done is, is I took a recent study by the ADA on AI and technology inside of dental offices, and it's a long article. There's a lot of text, graphs, things like that. What I did is I used. I used Notebook lm, which is Google's. It's a very, very powerful tool from Google, where I was able to take that PDF, drop it into NotebookLM and ask it to create a podcast episode where two people are discussing the results, or better said, summarizing the explanations within that article. And I think you'll find it helpful. I'm not going to be using NotebookLM to run all my podcasts, but I do want to use this as an example to show you the power of AI in so many spaces of our lives and in so many spaces of our business as well. So I hope you enjoy the episode.

You know, think about all the tech that helps us out day to day. Like the GPS in your car.

Yeah. Guiding you around. Or even just, you know, voice commands on your phone.

Exactly. These, these digital helpers, they're getting incredibly sophisticated, often using artificial intelligence, AI or augmented intelligence. A ui.

That's right. And these aren't just, you know, concepts anymore. They're really starting to make waves in.

Healthcare, including right there in the dentist's chair.

Absolutely. Dentistry is seeing a lot of change because of AI. And yes, aui, it's happening now, both clinically and even, like on the admin side.

Okay, so for this deep dive, we're looking at a specific source, right?

We are. It's a really useful white paper from the American Dental Association's Standards Committee on dental informatics. The SCDI came out in 2022.

Overview of artificial and Augmented Intelligence Uses in Dentistry. Sounds like the perfect guide.

It is. So our goal today is really? To unpack what this paper tells us. We want to get a clear picture of where AI and AUI stand in.

Dentistry right now, how they're actually being used.

Exactly. And where things might be headed. And you know, what it all means for you, whether you're working in dentistry or you're the patient.

Let's start with those terms then. AI versus iu. The paper makes a distinction.

It does, and it's an important one. AI, Artificial intelligence. That's sort of the broad term. Machines showing intelligence may be mimicking human thinking.

Like a computer playing chess.

Kinda. Yeah. But in dentistry, the emphasis in this paper is really more on augmented intelligence. AI.

Okay, so how's that different?

Well, IU isn't about replacing the dentist. Not at all. It's about systems built specifically to enhance what the human clinician does.

Ah, so it's more like a tool.

Precisely. A very smart tool designed to assist the dentist, but keeping the human expertise, human decision making, front and center.

Got it. And underpinning both AI and AI of the algorithms.

Yep. The paper basically defines them as sequences of instructions. You know, rules a computer follows to solve something or process data.

Even the learning systems, machine learning, deep learning. Those terms pop up too, right?

Those are just different techniques, different ways. These complex algorithms learn from data, find patterns, especially in dentistry, analyzing lots and lots of images.

But the key takeaway, you said is that these are tools.

Absolutely. The paper stresses this again and again. They are designed to support people, especially clinicians, making those crucial care decisions.

So the dentist is always in charge. The final call rests with the human.

That's the principle. These systems use their algorithms, process the data, a ton of imaging data, to offer insights, maybe flag things, but they don't diagnose on their own.

Okay, that's a really important distinction. So let's get into the clinical applications. How is this actually showing up at the dentist's office?

Well, as the paper points out, a massive area is image analysis. Your X rays, 3D scans, photos taken in the mouth. AI is being trained to interpret those.

So what does that look like in practice? Can we walk through some examples?

Sure. Let's start broad. Just basic image interpretation, helping identify normal anatomy. Maybe flagging areas that look unusual or could be signs of disease.

Are you looking to buy or sell a dental practice? If you're a seller, how do you find a strong list of potential buyers? There's no MLS or Zillow for dental practice sales. In such a fragmented market, many dentists selling their practice feel discouraged. That's why I built practiceOrbit.com practice orbit is modernizing how dental practices are bought and sold. It's a platform technology. Y' all know them Zillow, Airbnb, Uber, and OpenTable, just to name a few. These technologies do two things. First, they help match buyers and sellers by centralizing the marketplace. Second, once the match between buyer and seller is made, these technologies help them navigate the sale to a successful close. Here are a few of the features inside a Practice Orbit users can determine a reasonable sale price for practice in just minutes. All they need is the most recent business tax return. Practice owners can use the directory to find important service providers like practice brokers, attorneys, accountants and bankers. Practice owners and their brokers can invite those service providers into the platform. So all parties are coordinating the sale together. Practice owners will have a beautiful digital listing of their practice, like a home on Redfin. And importantly, the listing will be anonymous and personal information like your address, name and pictures are only released after an NDA is signed and approved by you or your broker. Practice buyers can create saved searches to be auto notified when new listings come on the platform that fit their desired specialty and location. And hey, dental practice brokers, you can list all your practices on the site for free. By doing so, you'll get access to Practice Orbit's large pool of buyers, all of whom have created profiles and saved searches to be auto notified when new listings come for sale. You can use the platform to introduce your listings to a targeted set of buyers. Additionally, and optionally, you can use the platform tools such as a built in NDA, a customizable loi, and a workflow tool navigate the sale from LOI to closing. In fact, you can manage all your listings on the system, from listing the practice to NDA, to loi, and finally to closing the deal. Create your free account today at www.practiceorbit.com.

Just a first pass, you know, and.

Then thinking about prevention.

Are you looking to buy or sell a dental practice? If you're a seller, how do you find a strong list of potential buyers? There's no MLS or Zillow for dental practice sales. In such a fragmented market, many dentists selling their practice feel discouraged. That's why I built practiceOrbit.com practice orbit is modernizing how dental practices are bought and sold. It's a platform technology. Y' all know them Zillow, Airbnb, Uber, and OpenTable, just to name a few. These technologies do two things. First, they help match buyers and sell sellers by centralizing the marketplace. Second, once the match between buyer and seller is made. These technologies help them navigate the sale to a successful close. Here are a few of the features inside a Practice Orbit users can determine a reasonable sale price for practice in just minutes. All they need is the most recent business tax return. Practice owners can use the directory to find important service providers like practice brokers, attorneys, accountants and bankers. Practice owners and their brokers can invite those service providers into the platform so all parties are coordinating the sale together. Practice owners will have a beautiful digital listing of their practice like a home on Redfin. And importantly, the listing will be anonymous and personal information like your address, name and pictures are only released after an NDA is signed and approved by you or your broker. Practice buyers can create saved searches to be auto notified when new listings come on the platform that fit their desired specialty and location. And hey, dental practice brokers, you can list all your practices on the site for free. By doing so, you'll get access to Practice Orbit's large pool of buyers, all of whom have created profiles and saved searches to be auto notified when new listings come for sale. You can use the platform to introduce your listings to a target targeted set of buyers. Additionally and optionally, you can use the platform tools such as a built in NDA, a customizable loi, and a workflow tool to navigate the sale from LOI to closing. In fact, you can manage all your listings on the system, from listing the practice to NDA to loi, and finally to closing the deal. Create your free account today at www.practiceorbit.com.

Keeping small problems from Becoming Big ones?

Exactly. This is where it gets really interesting. AI systems could help spot issues much earlier. Think about evaluating images over time, like.

Comparing this year's X ray to last year's.

Right, AI might detect very subtle changes indicating early gum disease, periodontal issues. Or maybe spot things like congenitally missing teeth even sooner. Helping the dentist intervene earlier the paper.

Mentioned helping plan preventative treatments too.

Yes, like preventative resin restorations. And not just finding cavities, but maybe even quantifying them on an X ray, suggesting treatment options based on what it sees.

Okay, so Carrie's detection finding cavities seems like a prime example.

It really is. It shows how UGUI enhances rather than replaces. The dentist still uses their mirror, their explorer, or their judgment.

But the AI gives them another layer of information.

Precisely. It might improve detection, especially for those tricky early enamel carries. The paper cites one study showing AI.

Boosted sensitivity there, and you mentioned quantification.

Yeah, imagine instead of just looks like it might be getting better, the AI could offer a more objective measure of whether a spot is remineralizing based on analyzing the image density.

That's fascinating. And maybe even cleaning up the image itself.

Potentially, yeah. Auto correcting image quality issues to make diagnosis clearer. The bottom line for you as a patient, finding things earlier may be leading to less invasive, less costly treatment.

Okay, let's shift to gums and bone health. Periodontics, Right.

Measuring things like pocket depths around teeth or bone levels on X rays. Traditionally, there can be some variability doing.

That manually, so AI helps standardize that.

It aims to. By analyzing the images, AI can potentially improve the comprehensiveness, the accuracy, maybe the efficiency of collecting that crucial data.

Like measuring bone levels automatically.

Exactly. The paper talks about deep learning models measuring bone levels relative to the cej, the cementum amyl junction on radiographs. It could make assessments more reliable, more data driven.

Still challenges with very early bone loss though.

Yes, the paper notes that's still tricky, but overall, it's about getting a clearer, more consistent picture of periodontal health.

How about dental implants? Seems like a very visual area.

It is. And AI is being used there too. Models are being trained to recognize different implant types just from an X ray.

How accurate is that?

The paper cites ranges like 93.8% up to 98% accuracy in some studies. Pretty good.

Wow. And beyond just IDing the implant, predicting.

Success, training systems to estimate the likelihood of osteointegration, how well the bone will fuse to the implant.

Really? What kind of accuracy are we talking there?

The figures cited are between about 62% and 80.5%. So still developing, but promising. Also helping optimize implant design itself. Or detecting problems like implant fractures in.

Images, which for a patient could mean more predictable outcomes.

That's the goal. Yeah.

Okay, what about oral and maxillofacial surgery? That covers a lot.

It does. And the paper suggests while some uses are limited now, the future potential is huge.

Such as?

Well, things like predicting tmj, jaw joint internal derangements from imaging with pretty high specificity and sensitivity reported.

So helping diagnose jaw problems.

Right. Or assisting in diagnosing cysts, tumors, maybe even aiding early oral cancer detection by analyzing images for suspicious patterns.

That could be incredibly impactful.

Absolutely. And think about planning complex surgeries like orthognathic surgery to realign jaws. AI could potentially help with more precise diagnosis, choosing the best surgical approach, even predicting healing outcomes.

Any other examples?

The paper mentions an operation Smile Microsoft tool using AI to grade cleft lip surgery outcomes. And robotic surgery is emerging, though the Paper focuses more on the image analysis support side for surgeons.

It sounds like it's aiming for greater precision in really complex situations.

That's the idea. Reducing errors, improving planning.

What about endodontics?

Root canals, Another area where imaging is key. Especially complex 3D scans like CBCT and AI is seen as a valuable support tool here.

Helping the dentist interpret those images exactly.

Helping detect things like periapical lesions, infections at the root tip, or maybe tiny root fractures.

How reliable is that?

Studies cited show promise like 92.8% reliability for spotting those lesions on CBCT in one case. Also assisting with planning.

Like figuring out the exact length needed for the root canal filling.

Yeah. Determining working length, assessing complex root shapes, evaluating the quality of previous root canals. One study even found AI was more accurate than a human expert in locating the very tip of the canal, the minor apical constriction.

So aiming for more accurate diagnosis and really meticulous treatment planning.

That's the goal for better outcomes.

Now, you mentioned dental labs, prosthetics. That area is already quite digital, isn't it?

It is. Dental labs have been early adopters of digital tech, including elements that lean on.

AI aui Like designing crowns and britches.

Right. The paper talks about digital wax ups, AI helping design restorations, surgical guides for implants, even night guards or occlusal guides, all from digital scans.

Faster and more consistent?

Generally, yes. Improved speed, reproducibility. Labs also use it internally to streamline workflow. Things like automated scoring of cases, routing work, even automatically marking the margin line on a digital tooth preparation.

And for the patient, that means potentially.

Faster turnaround times for high quality, well fitting crowns, bridges or appliances.

Okay, zooming out a bit. Radiology and imaging in general. This seems central to almost everything we've discussed.

It really is fundamental. AIUI tools are touching almost every part of the imaging process in dentistry.

Like deciding which X ray to take.

Yeah. Helping select the appropriate exam. Then identifying and interpreting what scene carries bone loss, t TMJ issues and more advanced stuff. Using techniques like radiomics to find subtle patterns that might predict outcomes. Screening for other health issues visible on dental X rays. Like signs of osteoporosis.

Wow. And annotating things.

Yeah. Marking structures like nerve canals or jawbone boundaries, automatically enhancing image quality. Even analyzing the technical quality of the X ray itself.

The paper mentioned something about illuminating images.

Right? Like overlaying a standard black and white X ray with color coded areas or quantified data points. That helps the dentist. Sure. But it can also make it much clearer for you. The patient to understand what's going on.

Better communication, better understanding. Okay, how about orthodontics, braces and aligners?

Big impact there too. Partly driven by all the digital scanning and the rise of remote monitoring.

So how is AI used?

Assessing growth is a big one. Analyzing things like cervical vertebrae maturation on X rays to time treatment. Or predicting dental age. The paper cites pretty good accuracy figures, often more consistent than manual methods.

And treatment planning.

Absolutely. Cephalometric analysis, measuring skull and jaw relationships is being automated using deep learning. The paper mentions accuracy over 90% and importantly, higher consistency than humans doing it manually.

That sounds significant.

It is. Faster, more reliable analysis for diagnosis and planning. Plus other uses like facial analysis from scans, helping design treatment, even automating aspects of retainer construction.

So potentially more efficient and personalized ortho care.

That seems to be the direction.

And finally, the paper touched on TMD and sleep disorders.

Yeah. Tying imaging and AI together, helping identify airway structures, potential obstructions related to sleep.

Apnea and the jaw joint prediction. Again, right.

Predicting those TMJ internal derangements. Also like we saw with prosthetics, AI is being developed to help design custom night guards directly from intraoral scans for better fit, easier replacement.

So new tools for tackling those often tricky jaw joint and sleep breathing issues.

Exactly. Valuable additions to the toolkit.

Okay, so beyond these specific specialties, how is AI integrating with the sort of digital backbone of the practice, Teledentistry records scanning?

Good question. These technologies are becoming crucial data feeds for AI teledentistry.

For example, remote consultations.

Right. AIUI can help screen patients remotely. Maybe triage cases based on photos they send in. Or even monitor treatment progress by analyzing images over time.

But image quality could be an issue there. Someone's taking a picture with their phone.

That's a key challenge. The paper notes. Poor quality images could lead to false positives or negatives. It really underlines why the human dentist's judgment remains critical. They need to understand those limitations.

What about Electronic dental records?

AI can analyze the vast amounts of data in EDRs. Cross reference notes with images, maybe flag a potential misdiagnosis from a past visit. Or identify treatment that was planned but never scheduled.

And it needs good data to learn from, right?

Absolutely. The paper mentions efforts like the Big Mouth Database pooling de identified data from dental schools to train these models. Standardized, diverse data is key. And analyzing all that data helps personalize risk assessment. Your individual risk risk for caries or gum disease, for instance.

And scanning intraoral scanners seem much more common.

Now very common. The ADA survey mentioned in the paper found about half of dentists using them. These scans are vital inputs for AI.

Feeding the design process for crowns and things.

Yes, but also AI can help identify potential errors in the scan itself, recognize anatomical landmarks automatically, and just generally facilitate those digital workflows for making restorations, guides, aligners, you name it.

Okay, so that covers the clinical side and how it connects to practice tech. But the paper also talks about non clinical uses like the business side.

It does. And this is an area where AI UI is already quite active. Often used by third parties like dental insurance companies, the payers and clearinghouses.

How are they using it?

A major application is in claim processing. When a dental office submits a claim, the.

With X rays and documents.

Exactly. AIUI systems can automatically validate that the right attachments are there. And critically, they can screen the radiographic images.

They screen them for what?

To check if the treatment being billed or proposed seems to align with the payer's clinical guidelines based on the X ray evidence.

Okay, and how is that different from before scale?

A human consultant at the insurance company could only ever review a small fraction of submitted images. The paper highlights that AI can potentially review every single image. Wow. Yeah. So it can efficiently flag claims that look like they need a closer look. Routing just those to the human reviewers, it's a huge efficiency gain for them.

And that connects to payment integrity preventing fraud.

It does. The paper describes using computer vision to create a sort of digital fingerprint for.

Each image to see if it's been tampered with.

Right. Detect manipulation, cropping, brightness changes, things like that. Or spot suspicious patterns. Like the exact same X ray being submitted for multiple different patients.

Okay, that's clever.

It also assists in post payment audits, checking if the work is actually done by comparing post operative X rays to the services billed. Ultimately, this helps ensure dental benefits are used appropriately.

Are payers using it internally too?

Yes. The paper mentions development of AI UI for quality assurance. Basically monitoring the consistency of their own human dental consultants against standards or peers.

And are dental practices themselves getting these kinds of tools?

They're starting to. AIUI tools are being developed for use in the dental office to validate claims before submission.

Like a pre check?

Exactly. Checking if all the required documents are there, if the X ray supports the code being billed. The idea is to streamline that whole revenue cycle, reduce denials, and hopefully make the process smoother for everyone, including the patient.

Okay, it's clear this tech is becoming deeply embedded. Which brings us to the really important questions around Rules, regulations, validation. How do we make sure this is safe and effective?

Absolutely crucial, because again, the paper constantly reminds us AIUI is a supplement, a tool. The licensed dentist is responsible for patient care. Period.

Are there global guidelines?

There are efforts, yeah. The WHO has ethical guidance. There's the imdrf.

Do you ever feel lost in your practice finances? And how about your personal finances? Ever wonder where your hard earned money goes and why you're not seeing the financial progress you've worked so hard for? If so, listen up. As a dental CPA and financial planner, I've created our company Practice CFO just for you. With almost two decades of experience, Practice CFO integrates great CPA services like accounting and tax with customized financial planning, both business and personal. Think of it this way. Our dental focused CPA services give us the X rays we need to clinically plan your global finances and be effective financial planners. It's a highly customized approach we call the CFO Model. Historically, having a CFO or Chief Financial Officer was an advantage of large corporations only. But today Practice CFO gives that same advantage to private dental practice business owners. As a client of Practice cfo, you'll be assigned one of our dedicated CFO advisors. You'll receive insightful financial reports each month, have regular financial review meetings, reduce taxes, grow your wealth and feel more in charge of your numbers. Because we believe that in the competitive landscape of dental practice ownership today, you have to be an effective business owner, not just a capable clinician. Reach out to learn More More and www.practicecfo.com.

Work for software as a medical device. These tend to be advisory setting principles.

But in the US the FDA has a direct role.

Yes, software intended for medical purposes, including diagnosis or treatment support, what they call SAMD is regulated as a medical device. It requires pre market authorization based on risk. Exactly. Class 1, 2, 3, depending on the potential risk to the patient. And critically, the paper points out that risk level depends heavily on the intended use.

Can you give an example?

Sure. An AI tool that flags potential quality issues on an X ray for internal review might be lower risk class one or two maybe. But an AI tool intended to be used chairside to actively help diagnose caries which directly influences treatment decisions. That's likely higher risk class two or even three requiring much more rigorous evidence and control.

That makes sense. What about transparency? Knowing how these things work?

Big topic. The FDA has held workshops and promotes good machine learning practices. These are principles.

Like what things?

Like ensuring the data used to train and test the AI represents the intended patient population keeping training and testing data separate and providing clear information to the users, the dentists about what the system does, how it performs and its limitations.

And what about bias? That's a huge concern with AI. Generally it is.

The worry is that algorithms trained on data that isn't diverse enough could perform differently or unfairly for certain groups based on race, ethnicity, gender, etc.

Are regulators addressing that?

Yes. The paper mentions efforts by agencies like AHRQ and advisories from the FTC urging companies to actively test for and mitigate bias and to be transparent about it. Making sure these tools work equitably for everyone is vital.

And patients getting access to this information.

That ties into the 21st Century Cures Act. It mandates easier patient access to their digital health info, which could down the line include reports or assessments generated by AI based on their images or data.

The paper also mentioned the White House AI Bill of Rights.

Yeah, just noting that principles are meant to guide responsible AI deployment across sectors, including healthcare.

Okay, technically, how do they prove these systems actually work?

Validating, right? You have to to test the AI system against independent data where you already know the right answer, the ground truth. Usually this involves experts carefully reviewing and tagging images or data.

Do you ever feel lost in your practice finances? And how about your personal finances? Ever wonder where your hard earned money goes and why you're not seeing the financial progress you've worked so hard for? If so, listen up. As a dental CPA and finance financial planner, I've created our company Practice CFO just for you. With almost two decades of experience, Practice CFO integrates great CPA services like accounting and tax with customized financial planning, both business and personal. Think of it this way. Our dental focused CPA services give us the X rays we need to clinically plan your global finances and be effective financial planning planners. It's a highly customized approach we call the CFO model. Historically, having a CFO or Chief financial Officer was an advantage of large corporations only. But today Practice CFO gives that same advantage to private dental practice owners. As a client of Practice cfo, you'll be assigned one of our dedicated CFO advisors. You'll receive insightful financial reports each month, have regular regular financial review meetings, reduce taxes, grow your wealth and feel more in charge of your numbers. Because we believe that in the competitive landscape of dental practice ownership today, you have to be an effective business owner, not just a capable clinician. Reach out to learn more at www.practicecfo.com.

But the picker raised some concerns there.

It did. A key challenge noted Is that the validation data sets used by manufacturers might sometimes lack transparency or might not be rob enough, not diverse enough, maybe inconsistent ground truth tagging. There is a call for better standards here.

And that comes back to the dentist again, doesn't it? Responsibility?

Absolutely. Clinicians need to understand the tool, but not blindly trust it or over rely on it. They need to know its performance characteristics.

Like how often does it get things wrong? False positives, false negatives.

Exactly. And that performance might differ depending on how it's being used. Just for screening versus making a final treatment decision. The paper suggests dentists should ask vendors tough questions. What was this validated for? What kind of data was used? What are the expected error rates?

So pulling this all together, what's the impact on how a dental office runs the workflow? And more importantly, what are the benefits for you as a patient now and maybe down the road?

Well, workflow wise, AIAUI acts like that super assistant, preparing data, processing images, highlighting things for the dentist. It can make their process more efficient, potentially freeing up time and looking further ahead. The paper paints a picture of a future where this could lead to a more integrated oral systemic health system using predictive analytics, maybe even achieving precision oral medicine.

Tailoring treatment based on vast amounts of data.

That's the vision. Though the paper cautions ensuring those predictions are fair and unbiased is still a hurdle.

Okay, but for the patient sitting in the chair today or soon, what are the potential upsides?

Some are maybe indirect efficiencies that could eventually help manage costs or faster insurance processing. But more direct benefits, enhanced diagnosis, hopefully catching problems earlier. Maybe more accurately, this could lead to better treatment choices, better material selection, and by analyzing trends in your data over time, better monitoring of your oral health.

The paper mentions specific things like calibration.

Yeah, getting more uniform standardized assessments like a consistent score for care's risk or periodontal health, regardless of which clinician you.

See and understanding things better.

Right, that visualization aspect, seeing the AI highlight things on your X ray can really improve your understanding and maybe even trust in the diagnostic process. Plus general efficiencies in the practice should ideally translate to a smoother, better experience for you.

So the overall message from the paper seems to be AI and AUI are here. They're tools.

And if used correctly, if adopted thoughtfully, safely, with good validation and oversight, they hold significant promise for improving dentistry.

We've definitely covered a lot of ground in this deep dive from how AI AUI is touching almost every dental specialty to the behind the scenes admin uses and the crucial regulatory pieces, all based on that ADA white paper.

Yeah, it's clear this isn't just theory. This tech is integrating rapidly, acting primarily as that powerful tool to augment what dental professionals do. The potential for gains in efficiency, accuracy, and patient care is definitely there.

And hopefully understanding this gives you, our listener, a valuable perspective on the technology you might encounter in your own dental visits, and just how much thought goes into bringing these tools into practice responsibly.

So, as AI and AUI keep getting smarter and more integrated into dentistry, maybe here's a final thought to chew on. Given how fast this is moving and the complexities around validation bias and just understanding how these systems work, how do we ensure this powerful technology truly serves the best interests of every single patient? Balancing algorithmic insights with that irreplaceable human judgment and pair.