The symposium titled “Artificial Intelligence in Dentistry,” organized by the Üsküdar University Faculty of Dentistry, was held at the Ibn-i Sina Auditorium of Üsküdar University NP Health Campus.
The symposium, organized to academically address the transformative role of health technologies and artificial intelligence (AI) in dental diagnosis, treatment, and planning processes, discussed the scientific aspects of AI algorithms' potential to increase diagnostic accuracy, strengthen clinical decision support systems, and create personalized treatment plans in dentistry.
Prof. Ergün Yücel: “Artificial intelligence is a clinical practice today”
The program began with the opening speech of Prof. Ergün Yücel, Dean of the Üsküdar University Faculty of Dentistry. Prof. Ergün Yücel emphasized that artificial intelligence is no longer a future concept but has become part of daily practice in healthcare.
Stating that artificial intelligence has moved beyond being a theoretical discussion topic, Prof. Yücel said, “Artificial intelligence is not a future scenario today; it is now a clinical practice from our perspective. The rules of the game are being rewritten, so to speak, in all major branches of our profession, from diagnosis to treatment planning.”
The probability of error will be zero
Prof. Yücel stated, “For years, in dentistry, we said ‘first, do no harm.’ But today, with artificial intelligence technologies, we have moved a step further and are saying, ‘Reduce the possibility of error to zero.’ This might become a concept with artificial intelligence.”
Artificial intelligence is already being used
Stating that artificial intelligence is already actively used in many fields, Prof. Yücel said, “In radiology, we can capture details that might escape the human eye. In surgery, we use applications that guide with millimeter precision. We utilize these technologies as clinical support mechanisms from orthodontics to endodontics, pedodontics, and other areas.”
However, Prof. Yücel also emphasized that technology should not be sacralized, stating that artificial intelligence cannot replace the art of medicine, adding, “Even the most advanced algorithm cannot replace the bond of trust a physician builds with their patient and the human touch. No algorithm has empathy.”
Prof. Yücel concluded his speech by stating that artificial intelligence does not threaten the profession; on the contrary, it reduces the margin of error and provides advantages in time management, saying, “I hope this meeting will be a first and will be further developed with different applications in terms of scientific infrastructure.”
Prof. Türker Tekin Ergüzel: “The systems we develop do not replace the physician”
In the first session held after the opening, the general framework and ethical dimension of artificial intelligence in healthcare were discussed. Speaking at the session, Üsküdar University Vice Rector Prof. Türker Tekin Ergüzel shared real-world applications, particularly those based on clinical data, in his presentation titled “Artificial Intelligence and its Applications in Healthcare.”
Prof. Ergüzel stated that they conduct studies using clinical data generated at NPİSTANBUL Hastanesi, which has been operating within the university for approximately 15 years, and said, “All the systems we develop are preliminary diagnostic systems. These are decision support systems. The decision belongs to the physician; our systems support that decision.”
Artificial intelligence's greatest need is data
Emphasizing that big data is at the core of artificial intelligence applications, Prof. Ergüzel stated that with the Industry 4.0 process, there has been a shift from mass production logic to a “big data-oriented” structure, and said, “In daily life, you all generate data. Your phones recognize your face, recognize your voice. Behind this are signal processing and image processing algorithms.”
Prof. Ergüzel stated that with the transition from machine learning to deep learning, the volume of data has increased exponentially, leading to the involvement of GPU-equipped and neuromorphic computers. He explained, “A neuromorphic computer mimics a neuron. It processes data and acts as both temporary and permanent memory. It is incredibly fast.”
Ant, swarm, and bee algorithms
In his presentation, Prof. Ergüzel also included biomimetic (nature-inspired) optimization algorithms, explaining that ant colony, swarm intelligence, and bee algorithms are used in health data analysis. Citing the mechanism by which ants find the shortest path via pheromones, Prof. Ergüzel said that this logic is used to select the most meaningful features from data, adding, “We were achieving 60% accuracy with 48 features. When we used ant colony optimization, 22 features were selected, and accuracy increased to 80%.”
Emphasizing that swarm intelligence should not be confused with herd mentality, Prof.Ergüzel stated, “Herd mentality involves unquestioning imitation. Swarm intelligence, on the other hand, involves making rational decisions by utilizing the experiences of others.”
Prof. Ergüzel exemplified the bee algorithm through the “waggle dance,” stating that bees can provide the precise address of a nectar source up to 6 kilometers away.
Emotion recognition model with 40 thousand data points
Prof. Ergüzel stated that in addition to psychiatric diseases, they also conduct emotion recognition studies based on facial expressions, announcing that they have classified 7 basic emotions using over 40,000 data points.
At the end of his speech, Prof. Ergüzel drew attention to the responsibility of the education system alongside the opportunities offered by artificial intelligence, stating that young people should be trained in areas that machines cannot easily perform.
Prof. Bellaz: “Artificial intelligence in dentistry is not just a new device, it will change the paradigm”
Prof. İbrahim Berk Bellaz, a faculty member from the Department of Prosthetic Dentistry at Üsküdar University Faculty of Dentistry and a Prosthetic Dentistry Specialist at Üsküdar University Dental Hospital, evaluated the ethical boundaries and societal implications of artificial intelligence studies in his speech titled “Scientific Knowledge, Ethical Issues, and Societal Prejudices.” He stated that artificial intelligence signifies not merely a technological innovation, but a paradigm shift capable of transforming scientific thought and the identity of physicians.
Stating that perceiving artificial intelligence in dentistry merely as a new device is an incomplete approach, Prof. Bellaz posed the question, “How will we approach artificial intelligence decisions in an era where even scientific truth is not absolute?”
Artificial intelligence does not make decisions; it provides decision support
Prof. Bellaz stated that it is dangerous to view artificial intelligence as an “omniscient fortune-teller,” adding, “Artificial intelligence scans existing data, presents probabilities, and offers a decision support mechanism. It does not make the decision itself.”
Prof. Bellaz drew attention to the inevitability of ethical and legal discussions alongside artificial intelligence, raising the question, “Who will be responsible for decisions made in an era where scientific truth is not absolute? The physician, the software developer, or the algorithm?”
The physician's identity will change
Prof. Bellaz stated that with the introduction of autonomous robots, human-specific limitations such as physical fatigue and manipulation errors will decrease, and added, “The new physician will evolve into a profile focused on empathy, decision coordination, and preventive medicine.”
Prof. Bellaz stated that a mandatory “artificial intelligence formation training” for the use of artificial intelligence in diagnosis and treatment might also be on the agenda, saying, “Just as a formation training is required for teaching, a similar training might be essential for physicians using artificial intelligence.”
Prof. Yumuşhan Günay: “Artificial intelligence can be a compass in planning, but we are still at the beginning of the road in design”
Prof. Yumuşhan Günay, a faculty member from the Department of Prosthetic Dentistry at Üsküdar University Faculty of Dentistry, comprehensively addressed the role and limitations of artificial intelligence in implant planning with his presentation titled “The Use of Artificial Intelligence in the Production of Subperiosteal Implants” at the symposium.
Prof. Günay stated that he prepared his presentation with his thesis student and intern dentist Furkan Hastaoğlu, saying, “This topic is also our student’s thesis work. It serves as interim output from a research process we are conducting together.”
Classical implants cannot be applied in nearly 30 percent of the population
Emphasizing the importance of subperiosteal implants, Prof. Günay reminded that the prevailing concept in implant design today is the “dowel logic” approach. Stating that “there is a population of nearly 30% where these types of implants cannot be clinically applied,” Prof. Günay highlighted that cases requiring additional surgical intervention or still being unapplicable despite it constitute a significant patient group.
Strong in literature and planning, weak in three-dimensional design
Prof. Günay stated that artificial intelligence is strong in literature and planning but weak in three-dimensional design, adding, “There could be two main reasons for artificial intelligence’s weakness in the design phase. First, the poverty of the database. Second, the deficiency in visual processing capability.”
Prof. Günay emphasized that artificial intelligence provides “incredible savings and convenience” during the planning and literature review stages, but stated that it has not yet met clinical expectations when it comes to three-dimensional realistic anatomical modeling.
Prof. Günay stated that artificial intelligence should not be seen as a tool for shortcuts, saying, “Using artificial intelligence by equating our expectations with convenience would be a flawed approach. The control should remain in our hands. As of now, artificial intelligence is very powerful as an auxiliary diagnostic and planning tool, but it is not yet at a point where we can transfer the responsibility of design.”
What was done at the symposium?
In the second session, Asst. Prof. Fatma Aslı Konca Taşova discussed artificial intelligence applications in orthodontics, while Asst. Prof. Anıl Özgün Karatekin explained digital guidance, augmented reality, and robotic support in endodontics. Prof. Hacer Şahin Aydınyurt evaluated the process ranging from smart diagnostic systems to personalized treatment approaches in periodontology. Asst. Prof. Büşra Sınmaz conveyed the segmentation of temporomandibular joint (TMJ) structural components from MR images using deep learning methods in light of scientific data. In the final part of the symposium, Asst. Prof. Hazal Abat addressed the detection of pathological conditions in impacted third molar teeth using artificial intelligence models.Asst. Prof. Ece İrem Ravalı Ertan shared current developments regarding artificial intelligence applications in oral, dental, and maxillofacial surgery. The program's final presentation was delivered by Instructor Yaren Dilci Halmedow. Halmedow shared the future perspective on the use of artificial intelligence and smart agents in prosthetic dentistry with the participants. A group photo was also taken with the participants at the symposium.
