Is it as good as film? Wrong question.
Is it diagnostic?
What makes a digital radiograph diagnostic? It is a combination of image and software. Dentists tend to focus on the sensor and the final image but it is the software that matters more – or at least as much – as the sensor when it comes to delivering a diagnostic image.
In this regard I have long been a fan of XDR Radiography.
Doug Yoon, the founder of XDR is a dentist, a scientist and one of the smartest people I know in dentistry. Doug recently reminded me of the XDR motto “Imaging Through Science”. What that means in practice is that the team at XDR is constantly looking to refine the digital x-ray process through scientific investigation based on diagnostic efficacy. Sadly some vendors concentrate on what looks good and what will sell rather than on what is the best diagnostic result. Instead of a lot of cool looking but useless enhancements XDR actually uses advanced image enhancement to improve diagnostics…what a concept.
Don’t get me wrong, XDR has one of the best sensors on the market as well as great software.
Are you up for some in the weeds, future focused, nerdy stuff? If so keep reading.
The opening question about film refers to our early attempts to diagnose dental disease from a digital image by looking at it, just we used to do with film. But do we really need to look at an image to diagnose? Do we even need an image? No, we do not.
When we diagnose an x-ray image, we are using our eyes combined with knowledge, training and experience to detect changes in tissue density. These density changes are indicated by how opaque the image is. We have learned that lack of opacity indicates loss of density which indicates pathology. We can see changes in the image because the pixels are displaying different shades of gray. Here is the part we tend to ignore. The computer knows the difference in each pixel even if it does not display it as an image.
Imaging software can be trained to distinguish patterns and changes in grayscale that are too subtle for the human eye to detect. Once the software has been trained to know what pathology looks like whether it is interproximal caries, periapical pathology or crestal bone loss the software can render a diagnosis without rendering an image. That is what machine learning and artificial intelligence is all about.
We are a ways away from total digital diagnosis and will certainly pass through a significant hybrid step where the software does a preliminary diagnosis and also renders an image for the human to look at. However, once you get past the film-based paradigm from the early days of Roentgen you can see that we do not need a picture we just need a data steam that can be used to detect tissue changes that indicate pathology.
Until we get there we will still need to use our eyes, knowledge, training and experience looking at the best image we can get to diagnose dental disease. XDR will help you do that.
Dr. Stephen Abrams is a dentist in Toronto, he is also the creator of The Canary System a revolutionary system to detect changes in the crystalline structure of dental enamel. This allows us to detect early decay, enamel dysplasia and even early fractures which are undetectable using conventional diagnostic procedures. Recently he wrote me the following regarding dentistry in the pandemic and how his device, the Canary, can help.
COVID 19 really has had an impact on the dental profession and the entire dental industry. We are just now starting to see what the “new norm” will look like as dental offices open. In Canada, US and Europe there are a wide range of restrictions in place as dental offices start to re-open.
The critical issues for Dentistry are:
- Limiting or eliminating the production of aerosols during dental procedures,
- Integration of personal protective equipment and infection control into a dental practice workflow which impacts time and increases costs
- The impact COVID 19 has had on oral health
- Increased incidence of cracked or fractured teeth due to stress induced parafunction which can be detected by radiographs or other caries detection devices such as CariVu or SPECTRA
- Increased incidence of white spot / ICDAS 1 or 2 lesions due to altered diets and schedules at home
- Impact of having dental offices closed for up to 3 months in some states and provinces and then slower access to booking patient appointments
- Regulatory issues with aerosol production when using intra-oral radiographs
Dental regulatory guidelines are telling dentists to avoid intra-oral radiographs and to use panoramic or CT Scans instead. These types of radiographs will not provide any information on cracks or interproximal caries.
In terms of patient needs, in our clinical practice, our patients are contacting us regarding either abscessed teeth or fractured or broken teeth or restorations. The vast majority of the e mails from patients involve fractures. Our patients who have been diagnosed with TMJ issues and parafunction and have bite splints are fracturing teeth as well. This is no surprise as these are very stressful times. We are using The Canary System to examine the rest of the dentition to detect cracks and then developing a more comprehensive treatment plan.
We are also seeing a lot of white spot (ICDAS 1 or 2) lesions or early caries lesions when we begin seeing patients. From my discussions with our patient population, diet and regular oral hygiene has been quite a challenge especially with younger children.
The final issue is that the media and now our patients are asking about how and when we will be seeing patients. They are expecting changes in the office and I suspect may be reticent to have restorations placed.
The Canary System addresses these problems:
Accurate caries diagnostics especially around the margins of restorations
Accurate detection of cracks before the tooth fractures
The ability to measure changes in lesions over time – in response to various preventive / remineralization therapies such as Curodont Repair or Fluoride Varnish
No aerosol production since scanning can be done on teeth dried with cotton rolls or gauze
Canary System examinations are covered under American and Canadian Dental Association caries diagnostic codes
New revenue stream for the office which does not involve staff using extensive amounts of N 95 mask, gowns, face shields etc. since there is no aerosol production
Fitbit and other wearable devices typically linked to exercise are being studied as ways to identify people who are potentially infected with COVID-19 before symptoms appear, when they can unknowingly spread the disease.
The linked article hits many of my hot buttons regarding future tech in medicine and dentistry. It is also up to the second relevant as it provides a potential break through in Covid 19 diagnosis.
it seems that researches could identify people who had contracted Covid 19 based on changes in heart rate before any of the overt symptoms were apparent. Click the link for the details.
Reportedly researches could detect 80% of users who had Covid 19 based on wearable data. What the article does not state is how many false positives were detected. In other words did Fitbit data suggest a person had Covid who in fact did not? In many ways a large percentage of false positives is a bigger problem than the 20% who had it who did not receive the early warning diagnosis from their wearable.
The potential to improve the human condition with this type of technology is awesome. Would it not be fantastic if our health could be constantly monitored by that big data AI in the cloud and potential diseases detected and treated early? Wouldn’t it be wonderful if we knew we had early asymptomatic Covid and we could avoid infecting the rest of the family? Would it not be great if the doctor in sky knew everything about us and could remind (not nag…remind) us to get more sleep, loose a few pounds and skip that extra six pack of Bud?
Okay you see the problem. Who has access to all that extremely personal health information? Is it safe?
RELATED: Garmin a major supplier of GPS and other wearable devices has suffered a MAJOR cyber attack.
How do you find decay? Poke the tooth with a sharp stick? How do you pick a shade? Hold up a paint chip and pick what looks the best? How do you measure a perio pocket? Push in a dip stick? How do you check for TMJ problems? Put your fingers over the joint and listen for a click? Or is it a pop or is it grating or crepitus (whatever that is?)
All of these traditional diagnostic methods have several things in common; they are subjective, they vary significantly from one operator to the next, and they aren’t very accurate.
Digital hi tech diagnosis changes everything. The core of hi tech diagnostics is the ability to scientifically measure a biologic condition and then to convert those measurements to a digital format for analysis and recording. The result is a highly accurate and consistent measurement. It is similar to the difference between two baseball fans arguing over which pitcher looks to them to have the fastest fastball. Or taking out the radar gun and actually measuring.
From ScienceDirect: Following is the summary of this scientific article by E. Dianne Rekow, King’s College, London.
Digital innovations have unquestionably disrupted dentistry. With these innovations, patient experience has improved. More restoration options are available delivering longer lifetimes, and better esthetics. Fresh approaches are bringing greater efficiency and accuracy, capitalizing on the interest, capabilities, and skills of those involved. New ways for effective and efficient interprofessional and clinician-patient interactions have evolved. Data can be more efficiently mined for forensic and epidemiological uses. Students have fresh ways of learning. New, often unexpected, partnerships have formed bringing further disruption — and novel advantages.
Is digital dentistry disruptive? Absolutely. Is it destructive? Absolutely not!
I have been saying the same thing for many years. It is nice to see some research backing me up.
This story is not about…you know what. From UPI.com:
A new study casts doubt on claims that artificial intelligence (AI) equals or surpasses the ability of human experts to interpret medical images.
They claim earlier studies were flawed.
From Becker’s Review: If you read the article it doesn’t really amount to much. Supposedly no pathology was missed and no one was damaged by the failure to read the scans. If that is true, is there really any value to insisting that all CBCT scans be read by a trained radiologist?
At this time the pathologists and radiologists assert that all scans should be read by an expert and they make some good points as to why. However in my experience most scans are not being submitted for investigation by a specialist. Is that a breach of standard of care? or is it just lazy? Or is is actually OK if it makes no difference to the patients?
The best high tech solution will be for a trained AI to do a preliminary evaluation of all CBCTs and identify any that need additional evaluation.
Radiologists at Bethesda, Md.-based Walter Reed National Military Medical Center failed to read and diagnose nearly 2,000 dental scans, according to an internal investigation cited by The Wall Street Journal.
SmileSnap is the Leading Virtual Consultation Software and App for the Dental Industry. Grow your practice and start more patients with our patient analytics and metrics dashboard.
I just talked to the developers of this new service. I like a lot of the things they are doing. I am a huge fan of Teledentistry. At this time it is rather basic and is mostly a marketing rather than a diagnostic tool. Nevertheless it is a good start and a creative way to leverage the ubiquitous use of smart phones, increase the power of your web page and differentiate yourself from the competition.
Wearable blood-sugar monitors deliver round-the-clock glucose readings—and relief from the daily grind of finger-stick blood testsB
Diabetes patients are increasingly using electronic skin patches and their phones, instead of pricking their fingers, to do the complex job of managing a disease that affects more than 30 million Americans.
The so called healthcare crisis is not a crisis in care it is a crisis in affordability. This is exactly the way technology will reduce costs and solve the crisis.
From Study Finds:
…an artificial intelligence program trained to identify and diagnose skin cancer in humans proved to show greater accuracy when pitted against dermatologists.
Researchers from the U.S., Germany, and France used an AI system called the Convolutional Neural Network (CNN) to detect skin cancer by showing it more than 100,000 images of malignant melanomas and benign moles. They then compared the AI’s performance in successfully identifying them on human patients with the performance of 58 international dermatologists. They found that the CNN detected more melanomas and correctly identified more benign tumors than the dermatologists.
Another remarkable demonstration of the future of AI diagnostics. Question?
Would you feel better or even OK if you knew your suspicious skin lesson was examined by a machine and not by a human? The study tells us the machine is better, but how do you feel about that? Once we overcome our reluctance to trust the machine imagine this:
Joanna notices an unusual discoloration on her left shoulder. Concerned she uses her smart phone to take a photo and immediately uploads it to a diagnostic AI in the cloud. Within minutes she receives a message that the lesion she noticed is nothing to worry about. The cost to Joanna is almost zero in both time and money.
Compare that to our current system in which Joanna would need to call the dermatologist -wait on hold- make an appointment in a few weeks. Then she would need to take time off work or find child care drive to the doctor’s office, park, wait to be seen then have the doctor examine her. Once she had been seen and assured there was nothing to worry about the dance of the insurance coverage would begin. An office visit and skin exam will be charged out at least at $200, How much Joanna will need to pay as a co-pay or deductible is anyone’s guess.
Digital technology has the potential to drastically lower costs and improve results. The “crisis” in American healthcare is not a crisis of care it is a crisis of affordability. High tech diagnostics will solve this crisis much more effectively than a centralized command and control system.
From Los Angeles Times:
She told me the Health Insurance Portability and Accountability Act, aka HIPAA, the federal medical-privacy law passed in 1996, never anticipated a scenario in which people would willingly disclose confidential medical information to a private company via a voice-activated computer in their living room.
“HIPAA doesn’t apply to Amazon in this situation,” she said. “Amazon has no relationship with Alexa users as a healthcare provider. The company theoretically could do almost anything with that information.”
Very interesting on several levels.
First, what a wonderful service. You can simply ask Alexa a medical question and get an answer based on the best professional knowledge instantly. No calling for an appointment, taking time off work, finding a parking place, waiting, and waiting, then hassling with the insurer.
Second, The more we use the service the better Alexa will become at answering our questions.
BUT on another level what will Amazon do with all our intimate health related data? The bulk of the linked article is awfulizing, speculating on what Amazon might do.
I tend to be an optimist. The history of mankind is a relentless journey toward improving the human condition. Nevertheless I wish I trusted them more.
From Roll Call:
Imagine a not-too-distant future when medical devices powered by artificial intelligence continuously adapt to new symptoms presented by patients and learn how to make accurate diagnoses much like a well-trained physician would.
The Food and Drug Administration is preparing for such a future and weighing how to assess and certify such medical devices, seeing them more like living things that can’t be regulated in the same manner as old-fashioned equipment.
Interesting. I do believe a good case can be made in support of regulations to protect the public. On the other hand the bureaucratic incentive is to deny innovation. No one is ever criticized for denying a new device the public never hears about but they will be crucified if they approve some device that can be shown to cause harm.
It is reassuring to see there is at least some recognition of the problem.
This is so cool.
From GE Healthcare The Pulse:
In the Pirate Adventure, a visual transformation of the equipment that was available before, patients are on a dock. There is a shipwreck and some sand castles in the corner. Children then work on the plank to be scanned. The Coral City Adventure in the emergency room gives children an underwater experience. It has a disco ball that makes light like bubbles around the room; children get into a yellow submarine and listen to the sound of harps whilst the procedure takes place.
They have used creative 3M wraps to turn a scary MRI into a children’s fantasy ride. These are the same type of wraps used to advertise on cars and buses. The technicians add to the experience by making the procedure a fun adventure. The result is lower anxiety and much faster procedures, the technicians can do far more scans per day. This means not only is the experience better for the kids but it is more profitable for the center. Win win.
Why not do the same for pediatric dental centers? Turn the scary CBCT into an adventure.
Cone-beam CT (CBCT) performed better than periapical radiography in detecting the apical delta in more than 100 premolars, but both modalities fell short compared with the gold standard of microcomputed tomography (micro-CT), t
Cone-beam computed tomography (CBCT) is useful in detecting vertical root fractures (VRF) in vivo even when the fracture line itself could not be visualized on CBCT, according to Russian researchers
Source: Endodontics — April 12, 2019
What makes all of the amazing things possible we see with CBCT is that second “C” it stands for “Computed”. The fact that we have a digital image and it is analyzed by a computer allows us to see, or more accurately distinguish things that we could never actually see with our naked, unaided human eyes.
For example a single radiograph pixel could be one of at least 256 shades of gray. Our unaided eye cannot distinguish any difference between say shade 124 and shade 125. However the computer can “see” a difference. The computer can then be taught (programmed) to to find pixel patterns that indicate a pathologic change (such as a root fracture).
We often get hug up on what we can see on a displayed radiographic image and compare that to film. However the actual value of a digital image is not how it compares to film but that it is not film. A digital image contains far more diagnostic data than film, we just need to use the computer to help us see what is there.
Apple Watch ‘prolonged my life’ by spotting deadly heart illness just days after new ECG feature launched,
In a paper published on Monday in Nature Medicine, the scientists reported that they had built a system that automatically diagnoses common childhood conditions — from influenza to meningitis — after processing the patient’s symptoms, history, lab results and other clinical data.
The system was highly accurate, the researchers said,…
…Dr. Kang Zhang, chief of ophthalmic genetics at the University of California, San Diego, has built systems that can analyze eye scans for hemorrhages, lesions and other signs of diabetic blindness. Ideally, such systems would serve as a first line of defense, screening patients and pinpointing those who need further attention.
Another step to automated diagnosis.
The company’s mobile app enables people with type 1 or type 2 diabetes to track blood sugar, medications, meals and physical activity through their iPhone, according to the company. It also allows people to set daily goals and view progress as well as connect with a broader community and personal coaches.
Another interesting diagnostic that uses the power of a smart phone and leverages the communication and network capabilities of the Internet to improve treatment and results for patients with a chronic medical condition.