Bones and machines: AI and the transformation of dental implantology

Last month, Drs Jaesung Lee and Yuxiao Zhou, who are assistant professors in the departments of mechanical engineering and of industrial and systems engineering, respectively, received funding from the 2024 Seed Program for AI, Computing, and Data Science from the Texas A&M Institute of Data Science for their project titled “Toward smart orthopedic surgery planning by using physics-informed machine learning”. Their research seeks to create an AI model that is capable of rapidly assessing bone stiffness on CBCT images so that implant placement—and other orthopaedic surgery in future—can proceed with the greatest confidence. By integrating 3D image data, clinical information and biomechanical principles with advanced machine learning in the model, they aim to provide precise 3D pointwise estimations of bone stiffness and strain distribution.

It is important to ask, why is such an approach necessary and what existing techniques does it stand to supplant? Currently, bone stiffness is measured primarily in research settings using biomechanical testing, finite element analysis (a computer simulation method that predicts how materials respond to forces) or invasive indentation techniques, none of which are practical for routine clinical use. There is thus a wide berth for improvement, and like so many other spheres involving diagnostics and planning in dentistry, it is in AI that the answer is being sought for analysis of visual data.

Speaking recently to Dental Tribune International, Dr Lee explained the emerging importance of AI to orthopaedic surgery and implantology in particular: “AI-driven methods now allow surgeons to leverage detailed 3D imaging data for precise bone assessment, leading to highly customised surgical plans, reduced operative time and improved patient outcomes. Machine learning, particularly physics-informed machine learning (PIML), combines computational models with biomechanical principles to rapidly infer bone properties, enabling surgeons to make more accurate, data-driven decisions.”

He further elaborated upon what his and Dr Zhou’s research project seeks to do: “Our research aims to develop a precise, efficient and personalised method for dental surgery planning through innovative PIML. Specifically, we are focused on accurately predicting the stiffness and mechanical behaviour of bone tissue from CBCT images and patient-specific clinical data. By integrating these biomechanical insights with clinical patient factors using a machine learning framework, we intend to enable personalised predictions of bone response to surgical interventions. Ultimately, this will result in better-informed surgical decisions, reduced risk of dental implant failure and improved patient outcomes.”

As in other areas of dentistry being transformed by AI, the value of these computational systems is anchored in three simultaneous advancements: eliminating human error, providing instantaneous and precise analysis, and achieving these enhanced outcomes with reduced operational costs. This appears to be the established and inexorable path of progress for dentistry.

A further important benefit that such an approach stands to yield is that of the increasing personalisation of care. As AI-based technology affords an ever-greater degree of accuracy about each individual’s bone composition and properties, so too will dental implant treatment and outcomes become increasingly fine-tuned. Dr Lee emphasised these benefits: “The future of this field, particularly for dentistry, lies in increased personalisation and precision medicine through integrating AI-driven methods with clinical practice. Advancements in machine learning models will enable unprecedented accuracy in predicting patient-specific outcomes. In dentistry, we foresee a shift towards fully personalised care, where computational models guide every surgical decision—from implant selection and positioning to postoperative care. Additionally, as computational power grows, such personalised analyses will become routine clinical practices, significantly enhancing implant success rates, patient satisfaction and overall dental care quality.”

AI thus continues to transform, in overwhelmingly positive ways, the practice of dentistry. While it may be difficult for some to accept the partial displacement of human evaluation from the analytical sphere, use cases such as this, and many others like it, demonstrate unequivocally that these powerful computational models, when employed carefully, serve to enhance patient care.