Our website is currently undergoing maintenance. Article records and full texts remain available. Submissions can be made through the Editorial Manager.

Machine-Learning-Assisted Micro-CT and Multimodal Imaging for Osseointegration Assessment of Zirconia Ceramic Implants: Linking Processing, Microstructure, and In-Vivo Performance

Ya Qiu, Yang Liu, Ke Zheng

Journal of Ceramic Science and Technology, Vol. 17, No. 1 (2026), pp. 17-32

DOI: 10.4416/JCST2025-00027

Download full article (PDF)

Abstract

The transition from titanium to zirconia implants represents a paradigm shift in restorative dentistry, driven by aesthetic demands, biocompatibility concerns, and the need for enhanced long-term stability. This review critically evaluates the processing-microstructure-performance nexus of zirconia ceramics, emphasizing how sintering parameters, phase stability, and surface modifications collectively determine mechanical reliability and biological integration. The unique phenomenon of transformation toughening provides zirconia with competitive strength, yet susceptibility to low-temperature degradation, underscoring the importance of microstructural control and compositionally tailored systems such as ATZ and Ce-TZP composites. In vivo evidence reveals a heterogeneous landscape of survival rates, shaped by implant design and clinical application, highlighting both the promise and limitations of current systems. To address the shortcomings of conventional 2D histomorphometry, this review surveys the rise of high-resolution, non-destructive micro-computed tomography (micro-CT) and its integration with multimodal tools like Raman spectroscopy and nanoindentation for quantitative and qualitative assessment of osseointegration. The application of machine learning, particularly deep learning architectures such as U-Net and ResNet, is explored as a transformative solution for automated segmentation, morphometric analysis, and predictive modeling of clinical outcomes, with explainable AI (XAI) offering interpretability and trust. Finally, the convergence of advanced imaging and AI-enabled analytics is discussed within the framework of personalized medicine, where patient-specific digital twins may enable virtual testing and optimization of implant strategies. Collectively, these developments chart a pathway toward predictive, data-driven implantology and position zirconia as a viable and evolving alternative to titanium.

Keywords

Transformation toughening, low-temperature degradation, osseointegration assessment, deep learning, digital twin

View this issue · Article & DOI index