Estimated reading time: 5 minutes
Artificial intelligence (AI) is frequently referred to as the technology that will take humanity to a new frontier.
Yet while AI and machine learning have been around for some time, it’s only recently that the crucial role they play in helping to improve healthcare diagnosis accuracy, treatment and monitoring is being properly understood.
Even for an industry notoriously slow to adopt new technologies, the support it is able to offer oral health specialists is helping to transform the idea of the typical dental office and chair.
While it is still early days in terms of the application of this technology within the sector, AI and ML have the potential to be applied, and in some cases have already been developed, for a variety of uses including: the intelligent diagnosis of radiographs and photographs, insurance claim analysis, intelligent charting and treatment planning.
Used extensively in dental laboratories and now playing a growing role in dental education, it is even being used in dental offices as voice commands for taxing tasks, improving efficiency and also limiting contamination.
Distinguished Professor Fang Chen is an internationally recognised leader in AI and data science and the executive director of the UTS Data Science Institute.
Professor Chen says AI has been used as a diagnostic tool in many medical applications where it is coded based on human knowledge, before providing a standardised diagnosis based on the same principles.
“It can be set as a process to provide step-by-step diagnosis and give a well-shaped presentation. With AI-enabled tools, patient dependent variables can also be added, to facilitate better targeted and personalised care.”
The most obvious application of this is in the field of emergency dental care where the unpredictability resulting from the COVID-19 pandemic saw AI-supported software developed to allow patients to self-monitor while in their own homes.
Apps such as Pearlii, developed by Australian oral health specialist Dr Kyle Turner, see patients offered fast check-ups via their smartphones. The app works by getting users to take pictures of their oral problems using their smartphone which are then used to identify problems with teeth and gums.
Meanwhile, on the back of the uptake of smart toothbrushes—which connect wirelessly with a mobile app to track your brushing habits—University of Pennsylvania School of Dentistry staff recently developed a smart dental implant that resists bacterial growth. It even generates its own electricity through chewing and brushing to power a tissue-rejuvenating light.
Teeth restoration made easy
But it is in the orthodontic fields and radiology where AI is gaining the most traction.
In utilising genetic algorithms that aid in predicting sizes of unerupted teeth, virtual models and 3D scans are proving useful tools in assessing dental abnormalities and even craniofacial abnormalities.
Professor Chen says one of the key advantages to integrating AI and ML methods into prosthetic dentistry is that by using them alongside CAD/CAM to design crowns and bridges, it is helping to make restoring teeth a very “quick and easy” task.
This is because using AI methods, it’s possible to design onlays, inlays, crowns and bridges with greater accuracy, and design considerations can be customised to each particular case.
“As the entire process is facilitated by the computer, a tooth replacement can be perfectly measured, for example. AI can help to predict certain future changes to be considered, together with CAD/CAM to make suggestions for dental crowns and bridges,” she says.
Likewise, AI innovation group Pearl recently received authorisation to introduce an AI-powered pathology detection aid to Australian dentists to help in their evaluation of patient X-rays.
Called Second Opinion, the tool works by applying radiologic computer vision to highlight potential areas of interest, affording dentists a second set of eyes when reading X-rays.
One of the key challenges facing many Australian oral health clinicians is the issue of data retention.
And in this too, a solution can be found through the use of smart dentistry practices, says Professor Chen.
She says AI programs can provide compliance checks against requirements and also help to identify data gaps while performing certain predictive tasks.
Yet despite it being hailed as the great hope of dentistry worldwide, it is clear challenges still exist for the use of AI in dentistry—specifically in relation to data acquisition and ethical considerations.
No matter how intelligent, machine-driven dentistry cannot replace the human characteristics of clinical intuition and empathy in delivering personalised healthcare.
In a paper called the ‘Present and future of artificial intelligence in dentistry’, published in late 2020, Indian researchers suggested that while AI in healthcare had a very promising role, challenges “both in technical and ethical aspects” existed.
The paper noted that AI-based systems are machine based and controlled and conducted by computer scientists without any medical training which can lead to a problem oriented approach of AI application in healthcare delivery.
Nor can AI replace contemporary healthcare delivery models whose working completely depends on clinician skills and patient-clinician communication.
“A preferable suggestion is a model which accommodates both AI and human elements so that the process of data collection and categorisation becomes easy, and at the same time preserve the human aspects of clinical care,” their report noted.
For its part the Australian Dental Association says that AI and ML have the potential to lower barriers for timely and equitable access to oral healthcare, increase oral health awareness, and increase treatment compliance.
But it also says the application of AI and ML carries “distinct safety considerations” depending on stakeholder groups.
In its policy statement adopted in 2020, the ADA cautions that without appropriate transparency and governance, potential consequences of AI and ML misuse include adverse clinical, financial and/or reputational outcomes, data privacy and security breaches, and that these “could lead to loss of trust in healthcare professionals and organisations”.
It is a view echoed by Professor Chen who says the key ethical considerations of AI and ML use are around privacy and security of patients’ information, and transparency of AI decision-making.
“It should follow the ethical principles and practice guidelines on managing data privacy and security. It should also provide sufficient transparency to the users (such as dentists) on why certain suggestions are made, hence the dentists can decide whether they would like to trust and adopt AI suggestions.”