In many areas of the world, the population growth has outpaced the availability of clinicians, prolonging waiting times and placing greater demands on already full clinical schedules. Dental Tribune International spoke with Prof. Víctor Díaz-Flores García, an endodontist who also holds a law degree, about his research which indicates that artificial intelligence (AI) may be heralded as a technological hero to the rescue.
Prof. Díaz-Flores García, what is the focus of your research group at the European University of Madrid?
I, along with my co-workers Yolanda Freire Ana Suarez and Margarita Gómez Sánchez, am a part of the Transformative Research in AI and New Enhancements for Dentistry, or TRAINED, group. We are exploring the possibilities of AI in the field of dentistry, focusing particularly on patient safety.
The emergence of AI in the field of medicine as a whole is overwhelming, and it is necessary to study and understand how it can be useful not only for clinicians but more importantly for patients and the general population. Our group has published other papers on this topic, and our aim is to continue working in this area.
Were there any aspects of your findings that were unexpected? What do you feel are the key takeaways of your research as they apply to AI in dentistry going forward?
The use of ChatGPT, or any other app that uses large language models, must currently be understood as a tool in its infancy. Not all the information given to us is correct, so it is necessary to train the system to give increasingly correct answers. In the studies we have carried out with ChatGPT, we have seen a promising consistency in the responses to clinical protocols, but this must always be checked by the health professional when using these tools as a basis for diagnosis or treatment. It is obvious that in the near future these tools will be present in our daily treatment, but today we must understand that they are untrained systems that must be taken into account in their proper measure when making clinical decisions.
Your team’s publication indicated that AI could become an incredibly reliable assistant in oral surgery, particularly in areas of the world where dental specialisations are not yet recognised. What inspired you as researchers to explore AI’s capabilities in this direction?
Access to information is essential for decision-making. The availability of up-to-date protocols of clinical actions is a benefit for the patient and can avoid safety problems in many aspects (use of medicines, application of certain materials, etc). An AI program trained by professionals specialised in a specific branch of dentistry would be a valuable tool. As a working group, we have observed that there are protocols of great clinical value that can be very useful both for general dentists and for those who specialise in a particular field, so it is essential to have a tool that directly provides accurate information to questions posed to it.
Because large language models and really all AI programs rely on big data, what encouragement would you offer clinicians concerned about data usage?
States and international organisations are taking action through specific legislation in this area. As a group based in the EU, we are closely monitoring this legislation. Particular attention must be paid to the use of patients’ personal data when training an AI, as this data is sensitive and we could potentially be in breach of data protection laws and face serious legal consequences. Patients should also be aware of these issues and take care to avoid exposing their data to apps or systems that have nothing to do with genuine medical assistance, especially those that aim to commercialise such data.