Shaping the Future of the Clinical/Technical Relationship

A well-trained clinician and technician both respect the need for proper treatment planning, careful shade selection, quality impression taking, accuracy of fit, and successful final execution in the patient's mouth. Several technologies have emerged over the past year and will continue to improve as their AI (artificial intelligence) algorithms continuously learn and adapt with increased exposure to data points.

The basic premise of AI and machine learning starts and ends with predictability and reliability, two factors that clinicians and technicians strive for with each crown they prepare, impress, fabricate, and cement. Machine learning is a subset of AI technology that utilizes computational neural networks that learn from experience, similar to how human brains operate.^1,2 Just as the human brain ties together an exponential number of data points from experiences, AI networks attempt to replicate this transformative process and extract meaningful information from previously inaccessible data sources such as text, images, and videos. The beauty of AI and its application in dentistry is its ability to be highly accurate (in many cases, far more accurate than any human trained in that specific task) and far more reliable in the consistency of its outputs. AI algorithms have the ability to learn from both accomplishments and failures, similar to how a human brain reacts to both positive and negative feedback, and it is these attributes that have made AI technologies attractive to those in the dental industry.

Why is AI important and where can it be applied to help strengthen the clinical/technical relationship? Applications include but are not limited to:

1. Proper treatment planning, screening for disease, and allowing the clinician to feel confident that their prosthetic work is being built upon a sound and AI-verified foundation (Figure 1).

2. Impression taking and communication of marginal accuracy and preparation design for the prosthetic material of choice.

3. Fabrication of prosthetics utilizing algorithms to design esthetics based on facial shape, age, and a multitude of patient-specific characteristics.

4. AI validation of marginal fit during insertions and finalization of the case in the mouth (Figure 2).

5. AI-enhanced color and shade communication from clinician to technician (Figure 3 through Figure 5).

Several AI technologies are already available and are continuously learning and improving.

Proper treatment planning and clinical execution allow technicians to be successful with what they can deliver. Most often, technicians are at the mercy of clinicians with respect to the quality of data they receive and the treatment that has already been proposed and performed by the dentist. Advanced AI algorithms that are currently being rolled out to practices and laboratories will help overcome a major hurdle in treatment planning by taking into consideration facial esthetics, occlusion, vertical dimension, etc, to help automatically provide comprehensive treatment proposals for clinical cases that previously required an enormous amount of manual effort and experience.³ The ability to "design a smile" that will not only esthetically suit a patient but also function well long-term is a win for the clinician, technician, and patient. The use of AI algorithms to bring data points together from photographs, videos, and radiography allows a patient to look into the future of what could be possible, and many technologies are emerging to make this a reality. The challenge is that once a plan is decided upon, numerous manual tasks are required to make it a reality. AI tools will soon be available to assist in the quality of execution of treatment.

With respect to margin detection and definition, many CAD/CAM softwares and intraoral scanners have begun to utilize AI to better define margins and preparation design, and alert clinicians of problematic areas and the need for possible re-impression or preparation.⁴ Utilizing this technology at the source—at the time of impression while the patient is present—will positively impact the technician as they will be receiving more accurate and clinically acceptable data. Clinicians sending analog impressions for indirect restorations will have laboratory scanners utilize AI to automatically detect margins, or lack thereof, and inform the clinician that an impression may need to be re-taken in order to produce the quality result the clinician and patient are expecting.

When a clinician seats a crown in the mouth, they expect the marginal fit to be accurate to several microns, and algorithms built to detect marginal discrepancies will alert a clinician prior to cementation (Figure 2).

The ability to utilize AI algorithms to analyze tooth color, internal characteristics, and morphology from photography and 3D intraoral scans will serve as a true "game-changer" in the world of esthetics. One of the most challenging aspects of a master technician's job is matching a single-central incisor. Adequate shade selection requires copious amounts of skill, patience, and understanding from the technician, let alone the clinician, and many are at the mercy of their smartphone technology. A proper and often expensive camera, dual-flash setup, understanding of lighting, white balance, usage of a grey card, and polarizers are some of what goes into communicating shades effectively. Each step introduces inherent error and ultimately leads to very challenging situations that can result in remakes and patient dissatisfaction. The goal of the ceramist is to try to recreate nature. Enhancements in the AI interpretation of photographs, and the use of AI-enhanced spectrophotometry in-situ with the patient can allow for the entire tooth and both its intrinsic and extrinsic properties to be mapped and the data fed to a ceramic printer that would incorporate the incredible amount of detail that ceramists layer into their dental creation (Figure 3). Current attempts with multilayer blocks and pucks, and advancements in surface stains, are attempting to replace the laborious efforts of layering porcelain; however, the not-so-distant potential of printing ceramics could allow technicians to be more efficient with their time while still producing lifelike dental masterpieces for patients.^5,6 "For each patient, each tooth, I am trying to mimic nature with my porcelain," Master Dental Technician Yasu Kawabe says. "At the moment, the analog way of doing this is very difficult to surpass, but a computational understanding of color and 3D printing of ceramics in the near future may be the key."

It is safe to assume that the boom of AI in dentistry is only beginning and that its goal is to augment the work of clinicians and technicians and help them do what they already do well, better. The final result is longer-lasting outcomes, more prosthetic predictability, and ultimately a highly satisfied patient.

Christopher Balaban, DMD, MSc
Clinical Director,
Overjet AI
Clinical Faculty,
Boston University
Private Practice,
Boston, MA