Originally published in ADA on July 9, 2021
From diagnosis and treatment planning to practice management, augmented intelligence is positioned to transform the way dentists care for their patients.
“The adoption and implementation of augmented intelligence within the dental profession is poised to be a transformative technology in dental care delivery,” said Robert A. Faiella, D.M.D., chief dental officer of Overjet, which provides AI services to dental practices and insurance companies. “AI is just a technology, but if we create the desire for functionality, it is more than the technology alone; it connects us to a wider world of information and options.”
Augmented intelligence is the theory and development of computer systems that can perform tasks that would otherwise require human intelligence, such as visual perception, speech recognition, decision-making and translation between languages. The term may also be applied to any software that performs intelligent behavior and acts intelligently. Augmented intelligence is an extension of artificial intelligence in health care, emphasizing its assistive and supplementary role to health care professionals.
Dentists can use augmented intelligence in their practices in many different ways. Some applications include screening radiographs for bone loss, caries, calculus, crown indications and other findings; evaluating digital information such as radiographs, photographs and patients’ electronic health records to help make diagnoses and propose treatments; monitoring phone calls to improve patient communications; and making the insurance claim adjudication process more efficient.
“Dentistry is experiencing significant changes in the management and delivery of treatment associated with the extensive introduction of digital capabilities and informatics, augmented intelligence and machine learning, which is when systems are able to learn and adapt using algorithms and statistical models to draw inferences from patterns in data,” said Dr. Faiella, who is a past president and trustee of the American Dental Association. “The intent is to automate, streamline and improve the patient experience, the productivity of the practice and treatment protocols for the office.”
Where to start
Dentists who are interested in introducing AI into their practices should first start with a problem statement, said Amol Nirgudkar, co-founder and CEO of Patient Prism, a company that uses AI to analyze conversations between dental offices and new patients to help convert callers into booked patient appointments.
“For example, ‘I need more new patients. How do I get them?'” he said.
The answer could be for dentists to advertise their services, but how do they know if their marketing is working? That is where AI comes in.
“I could deploy AI to listen to phone calls that are driven by marketing and figure out whether I am scheduling those patients,” Mr. Nirgudkar said. “I could also leverage AI to understand whether marketing is bringing the right kind of patients into my practice.”
Another question could be, “Why is our unscheduled treatment percentage so high?” he said.
It could be because some issues are going undiagnosed or patients do not trust a diagnosis, Mr. Nirgudkar said. Applications that use computer vision — a field of computer science that focuses on creating digital systems that can process, analyze and interpret visual data, such as digital images and videos, at the pixel level — could help with those problems.
Understanding available AI applications and the potential development of additional models to advance the patient care journey also will help dentists looking to introduce AI, Dr. Faiella said.
“The introduction of new technology into a practice is always a concern to the clinician, primarily related to understanding impact, training and engagement in the workflow, and cost,” he said. “Tactics to onboard the new technology for the dental team, as well as the clinical workflow and data capture practices used to ensure AI insights, will build enthusiasm for the practice efficiencies to be gained not only for clinical implementation but also for administrative goals and practice productivity.”
Some dentists may have questions about the Health Insurance Portability and Accountability Act implications of using AI, but Dr. Faiella said the protection of patient health information should already be in place throughout the practice and by the vendors providing AI services.
Beyond human cognition
“The greatest impact of implementing AI is to gain insights beyond human capabilities across large datasets within the practice to identify disease occurrence, improve office efficiencies and advance care to our patients,” Dr. Faiella said.
Without AI, the knowledge and clinical expertise required to manage patients can be daunting, he said. Dentists and their staff must acquire information from patients’ medical and dental histories, interpret the possible influence of medical conditions and current medications, record examination metrics and radiographic imaging, and follow protocols to store and share the information securely.
David L. Botsko, D.M.D., a general dentist in Davenport, Iowa, and Ahmed Mahrous, B.D.S., a clinical assistant professor in the University of Iowa College of Dentistry and Dental Clinics Department of Prosthodontics, have developed digital teaching applications that give dental students experience with AI. They said the main strength of AI lies in its consistency.
“The biggest potential for AI in dentistry, as well as other health care entities, is in the standardizing of diagnosis/treatment planning, reduction in bias associated with radiographic analysis and diagnosis of tissue lesions,” they said.
The use of AI may also help patients trust diagnoses and treatment plans.
“Case acceptance has traditionally been based upon the trust relationship between the patient and provider. The use of technology to provide objective findings shifts the treatment acceptance from ‘trust’ alone to ‘trust with verification’ by demonstration of need,” Dr. Faiella said. “This not only allows the patient more control in the acceptance of treatment recommendations but also improves informed consent.”
AI can also improve the experience of patients in other ways. For example, Patient Prism uses natural language processing to understand phone conversations between new patients and dental offices. If a conversation does not result in a scheduled appointment, the company notifies the office, providing a visual depiction of what went right and wrong and presenting video training that addresses areas for improvement.
“A new patient’s first interaction with a dental office is mostly via phone. How they get treated on the phone defines the entire patient journey,” Mr. Nirgudkar said. “The strength of AI is to be able to analyze a missed opportunity quickly so that the dental practice can have a second chance to redo the first impression before the patient has already decided on another office or completely abandoned their quest to see a dentist. Theoretically, you could deploy humans to listen to all phone conversations, but it would not be financially or practically feasible.”
When it comes to insurance claims, the use of AI could make adjudication more efficient and objective.
“The use of AI-powered technology creates an opportunity for both the payer and provider to share common measurements when considering the documentation submitted for a claim review,” Dr. Faiella said. “Even with quantified review criteria, clinicians reviewing claims need to apply those standards consistently. And while insurers can reduce subjectivity and apply criteria more consistently at scale, providers utilizing AI-powered technology can apply the objective measures used by the insurers to minimize misinterpretation of clinical guidelines to speed claim acceptance and reduce the need for appeals.”
Room for growth
Mr. Nirgudkar sees opportunities for continued growth in the AI sector, including within the capabilities of his own software. The response time after a call started at three hours, and now it’s 15 minutes or less. He also sees the possibility for live AI coaching.
“AI will continue to get better and reduce the time to analyze and thereby allow greater chance of course correction,” he said. “In the future, the opportunity for growth here is to be able to have a live AI coach listening to conversations and guiding the receptionist in real time.”
He also thinks the use of computer vision in pediatric and orthodontic specialties will increase.
The applications developed by Drs. Botsko and Mahrous are helping to prepare the next generation of dentists to use AI.
One program allows dental students to propose treatment plans for different scenarios. Their proposals are then compared to ideal treatment plans based on the rules of treatment planning programmed into the application’s algorithm. The program is able to provide feedback based on how close the student gets to the ideal plan.
Another focuses on the design of removable partial dentures, again comparing student designs to the ideal designs based on the rules that have been programmed into the algorithm.
The ADA has formed a new standards working group to look at AI in dentistry. The group’s mission is to develop educational materials and determine best practices for the growing use of AI tools that support clinical decision-making in dentistry.
“AI and machine learning technology are here today to both improve our clinical intelligence in the delivery of care, as well as identify efficiencies for insurers and reduce subjectivity in the adjudication of claims to advance the delivery of care to our patients,” Dr. Faiella said.