Orthognathic and Cosmetic Surgery: The Frontier of AI, 3D Planning, Custom Implants, and Augmented Reality By Prof. Mohammad Sartawi
- Christanny Niatala
- Oct 23
- 8 min read
Introduction
In recent years, the field of maxillofacial surgery has entered a phase of rapid technological convergence. The union of artificial intelligence (AI), 3D planning and simulation, patient-specific (custom) implants, and augmented reality (AR) promises to reshape how we conceive, plan, and execute orthognathic and cosmetic facial surgery. In this article, I aim to survey the current state of the art (as of 2025), present representative studies and innovations, and discuss their clinical implications and future directions.
My objective is not to present my own study, but rather to offer a coherent and critical synthesis of the literature—so that colleagues at a congress or in professional media can understand where we stand today and what is on the horizon.
AI in Preoperative Planning: Capabilities and Challenges
Automated Segmentation, Landmark Detection & Plan Generation
One of the major entry points of AI into our domain has been the automation of image segmentation, anatomical landmark recognition, and proposal of virtual osteotomies/implant trajectories. In implantology, for instance, Che et al. (2025) reported a multicenter study of an AI‐based 3D implant planning tool, showing that the AI tool may reduce planning time and assist clinicians in standardizing trajectories, though discrepancies remained between AI and actual positions (i.e. further refinement is still needed) [Che SA et al., J Dent Res, 2025] (https://pubmed.ncbi.nlm.nih.gov/40876620/) (PubMed).
Similarly, a scoping review by Qiu et al. assessed how AI models perform in evaluating bone quality and quantity in implant planning from radiographic images, highlighting that AI can provide automated assessments (density, dimensions) that may support decision-making workflows [Qiu S et al., J Dent, 2025] (PubMed).
In another study, Roongruangsilp et al. (2025) compared two object detection models (Faster R-CNN and YOLOv7) across different implant planning software. They showed that both models achieved promising detection performance, with differences influenced by imaging rendering algorithms and software platforms. This underscores that integration and consistency across platforms remain nontrivial [Roongruangsilp P et al., BMC Oral Health, 2025] (BioMed Central).
These developments suggest that in orthognathic planning, AI could eventually help propose safe osteotomy lines or repositioning vectors tailored to each patient’s anatomy and aesthetic goals.
Limits: Validation, Generalizability, and Trust
However, the adoption of AI tools in clinical practice is contingent on several constraints:
Validation & robustness – many AI models are trained on limited datasets or from narrow populations; performance may degrade outside those distributions.
Standardization of metrics and protocols – reviews like Macrì et al. (2024) emphasize the lack of consistent protocols for training, ground truth labeling, and evaluation across studies in AI‐assisted implant planning [Macrì et al., MDPI Biomed, 2024] (MDPI).
Regulatory and interpretability issues – in surgery, the surgeon must remain in control; AI should serve as intelligible decision support, not opaque “black box.”
Integration with surgical workflows – capturing preoperative imaging, feeding data to AI, and converting outputs into usable surgical plans (guides, implants) must be seamless.
In the domain of orthognathic surgery, with greater bone movements and complex soft tissue interactions, rigorous validation in large series is still largely missing.
Custom Implants and 3D‐Printed Osteosynthesis Devices
How Custom Implants Are Designed & Manufactured
When proceeding from a virtual plan to physical execution, patient-specific implants (PSIs) have become an increasingly vital tool. The workflow generally involves:
Designing plates or segments in CAD (computer‐aided design) based on the planned segment positions.
Manufacturing via additive manufacturing (e.g. selective laser melting of titanium) to match the precise geometry.
Sterilization and intraoperative placement without bending or adjustment.
Compared to stock plates which require intraoperative bending and adaptation, PSIs reduce manipulation, improve fit, and potentially reduce operating time and errors of repositioning.
In facial reconstructive surgery, several series have attested to implant survival rates > 96% at 12 months and functional improvement (mastication, speech) when using 3D-printed devices [see reconstruction literature]—this gives confidence for their translation into orthognathic contexts.
In orthognathic settings, combining custom osteosynthesis plates with cutting guides ensures that the spatial relationships of the bones correspond exactly to the virtual plan, minimizing drift or cumulative error. Reported series in facial plastic and craniofacial surgery suggest “fidelity to plan” in over 90% of cases when using combined guides + custom fixation [Birbe Foraster J et al., Med Sci (Basel), 2025] (cited in earlier drafts).
Clinical Gains & Caveats
Advantages:
Reduced intraoperative adjustments: less time spent manipulating plates or re‐fixing segments
Improved accuracy: matches exactly the planned geometry, decreasing error margins
Potential for segmentation and modular implants: e.g. zygomatic, mandibular, or genioplasty components made separately
Challenges:
Cost & logistics: the design, manufacturing, and delivery pipeline can be expensive and time-consuming.
Lead time: for some cases, scheduling must account for implant fabrication lead time.
Regulatory acceptance & sterilization: ensuring compliance with medical device standards.
Stress shielding or mismatch: plate rigidity versus bone biology must be considered.
Overall, in well-selected cases, custom implants represent a practical step toward more predictable outcomes.
Augmented Reality (AR) and Navigation in Orthognathic Surgery
What AR Brings to the Table
Augmented reality permits the overlay of virtual plans onto the real anatomical field, offering the surgeon live guidance. In orthognathic surgery, this means seeing planned osteotomy lines, segment repositioning vectors, or screw trajectories projected onto the patient, rather than relying solely on static guides or navigation screens.
A hallmark study by Żelechowski et al. (2024) compared AR-based navigation techniques with conventional drilling guides in an orthognathic context. They tested two AR methods (one using ArUco marker tracking, another using infrared tracking) and found that AR navigation yielded comparable or even improved accuracy (mean deviations under 2 mm) relative to conventional guides [Żelechowski M et al., Healthc Technol Lett, 2024] (https://pmc.ncbi.nlm.nih.gov/articles/PMC11730987/) (PubMed). This work represents a paradigm shift demonstrating that AR is no longer a future concept—it is entering pilot clinical use.
Another relevant review, Advanced outcomes of mixed reality usage in orthognathic surgery (Stevanie et al., 2024), discusses the potential of combining virtual reality / AR with surgical simulation to enhance planning and intraoperative guidance [Stevanie C et al., J Korean Assoc Maxillofac Plast Reconstr Surg, 2024] (SpringerOpen).
Technical Considerations: Registration, Latency, and Tool Tracking
The success of AR guidance depends heavily on accurate registration (the alignment between patient anatomy and virtual plan). Errors in registration can propagate to misguidance. Some recent strategies involve using fiducial markers or markerless deep learning detection of landmarks to permit continuous image‐to‐patient alignment [Benmahdjoub et al., AO CMF blog, 2024] (https://www.aofoundation.org/cmf/about-aocmf/blog/2024_05-blog-benmahdjoub-augmented-reality-based-surgical-navigation) (aofoundation.org).
In addition, tool tracking is required. A promising approach is STTAR: using built-in cameras of AR head-mounted displays to detect retro-reflective markers on instruments without external tracking hardware. This method achieved translational accuracy of 0.09 ± 0.06 mm laterally, 0.42 ± 0.32 mm longitudinally, and rotational accuracy of 0.80 ± 0.39° in test settings [Martin-Gomez et al., STTAR arXiv, 2022] (https://arxiv.org/abs/2208.08880) (arXiv).
Latency is another issue: the lag between surgeon motion and AR update must be minimal (< tens of milliseconds) to permit real-time use. Hardware advances and optimized tracking pipelines are critical to reduce lag.
In sum, AR is approaching a maturity stage where it can be integrated into orthognathic workflows, but its reliability depends on robust registration, fast tracking, and well-designed user interfaces.
Synthesis: How These Innovations Work Together in Clinical Practice
When viewed in concert, these technologies form a pipeline:
Imaging + AI → preprocessing, segmentation, landmark detection
Virtual plan generation → osteotomies, repositioning, aesthetic simulation
Design of custom implants & guides → translating plan to physical objects
AR / navigation assistance → intraoperative execution, overlay of plan onto surgical field
Quality control and feedback loops → verification and potential intra-op adjustment
This integrated approach aims to:
Decrease cumulative spatial error (from imaging to final fixation)
Reduce operating time by eliminating intraoperative trial-and-error
Enhance symmetry and aesthetic predictability
Improve reproducibility across operators
Preliminary comparative series (in non-orthognathic settings) show that combining guides + AR may reduce deviations by 20–30% compared to guides alone. In orthognathic contexts, the AR-based method by Żelechowski achieved error < 2 mm, which is clinically acceptable in many jaw repositioning tasks [Żelechowski et al., 2024] (PubMed).
Yet, the extent of improvement in real-world craniofacial surgery remains to be systematically quantified in large patient cohorts.
Clinical Implications, Limitations, and Adoption Barriers
Benefits for Surgeons and Patients
Greater confidence in achieving the surgical plan
Shorter operative times and less fatigue
Better symmetry and aesthetic outcomes
Patient education advantage: visualizing plans preoperatively
Challenges and Risks
High costs of software, AR hardware, and custom implant production
Learning curve and training needs — surgeons must become comfortable with AR interfaces and AI outputs
Device and regulatory validation — not all AR systems or implants are certified
Registration drift or tracking failures — fallback strategies (e.g. revert to conventional guides) must be in place
Soft tissue behavior — these technologies mainly guide bone repositioning; prediction of soft tissue response remains less mature
Roadmap for Adoption
Pilot series and case reports in leading centers
Multicenter prospective registries to gather real-world outcomes
Development of open, interoperable software standards to allow cross‐platform workflows
Cost-effectiveness studies to demonstrate return on investment
Continuous refinement of AI models and AR hardware to reduce error margins < 1 mm
If the community can demonstrate consistent improvement over conventional methods, these technologies will no longer be optional—they will become foundational.
Conclusion
As of 2025, the combination of AI, 3D planning, custom implants, and augmented reality is not a distant vision but an emerging reality in maxillofacial and orthognathic surgery. While we are not yet at the point of fully autonomous navigation, the synergy of these tools promises gains in accuracy, efficiency, and aesthetic outcome.
In my view, the next 5 years will be decisive: the centers that adopt, validate, and refine these workflows will lead the field. My hope is that this article gives colleagues a clear snapshot of where we stand today—and what to watch for tomorrow.
References
Che SA, Yang BE, Park SY, On SW, et al. Clinical Evaluation of AI-Based Three-Dimensional Dental Implant Planning: A Multicenter Study. J Dent Res. 2025. PMID: 40876620. (full text via PubMed) (PubMed)
Qiu S, Yu X, Wu Y. Application of artificial intelligence in bone quality and quantity assessment for dental implant planning: A scoping review. J Dent. 2025 Aug 8:106027. doi:10.1016/j.jdent.2025.106027. PMID: 40784481 (PubMed)
Roongruangsilp P, Narkbuakaew W, Khongkhunthian P. Performance of two different artificial intelligence models in dental implant planning among four different implant planning software: a comparative study. BMC Oral Health. 2025;25:984. (BioMed Central)
Żelechowski M, Zubizarreta-Oteiza J, Karnam M, et al. Augmented Reality Navigation in Orthognathic Surgery: Comparative Analysis and a Paradigm Shift. Healthc Technol Lett. 2024;12(1):e12109. doi:10.1049/htl2.12109. PMCID: PMC11730987. (PubMed)
Stevanie C, Park J, Shim J, et al. Advanced outcomes of mixed reality usage in orthognathic surgery. J Korean Assoc Maxillofac Plast Reconstr Surg. 2024. (SpringerOpen)
Benmahdjoub M, Thabit A, Niessen WJ, Wolvius EB, Van Walsum T. Augmented reality based surgical navigation using head-mounted displays. AO CMF Blog. 2024. (on image-to-patient registration via marker detection) (aofoundation.org)
Martin-Gomez A, Li H, Song T, et al. STTAR: Surgical Tool Tracking using off-the-shelf Augmented Reality Head-Mounted Displays. arXiv. 2022. (marker detection accuracy: 0.09 ± 0.06 mm lateral, etc.) (arXiv)
About the Author
Prof. Mohammad Sartawi (Jordan) — Maxillofacial and cosmetic surgeon, Director of the Jordan British Specialty Center, and former President of the Jordanian Society of Oral and Maxillofacial Surgery. His lectures and expertise span orthognathics, rhinoplasty, otoplasty, and facial implants.
