Volume 2, Issue 1, Retrospective Study – Jan 10, 2026, Pages 71-78,

DOI: 10.64951/jmdnt.2026.1.5

Ayhan Yildirim¹, René Hertach², Vedat Yildirim¹

¹ Hochschule Zurich, Department of Medicine, Albisstrasse 80, 8038 Zurich, Switzerland
² Hochschule Zurich, Department of Dentistry, Albisstrasse 80, 8038 Zurich, Switzerland

Received: 26 June 2025, Revised: 05 November 2025, Accepted: 30 November 2025, Available online: 24 December 2025, Version of Record: 12 January 2026

© 2026 Journal of Medicine and Dentistry (JMDNT)

This article is published under the Creative Commons Attribution 4.0 International (CC BY 4.0) License.

You are free to share and adapt the material for any purpose, even commercially, as long as proper credit is given to the original author(s) and source.

Full license details

ABSTRACT

Background:
Artificial intelligence (AI) has shown promising results for retrospective outcome prediction in midfacial fractures. However, prospective evidence for its integration into clinical decision-making and patient counseling remains limited.

Objective:
To prospectively evaluate the feasibility, accuracy, and clinical impact of AI-assisted postoperative outcome prediction in midfacial fractures.

Methods:
A multicenter, prospective study was conducted in 2022–2024, enrolling 120 consecutive patients at Center A (Seeklinik Zürich, n=60) and Center B (Kieferchirurgie München, n=60). Preoperative imaging and planned surgical interventions were analyzed by the AI model, which predicted postoperative outcomes: enophthalmus ≥2 mm, malocclusion, reoperation, and overall complications. Surgeons were provided with AI predictions prior to surgery and documented whether these influenced surgical decisions or patient counseling. AI predictions were compared with actual postoperative outcomes using accuracy, sensitivity, specificity, and Cohen’s κ.

Results:
AI predictions demonstrated high accuracy: enophthalmus 92.5%, malocclusion 89.2%, reoperation 87.5%, and overall complications 90.0%. In 28% of cases, surgeons modified surgical approach or counseling based on AI predictions. Concordance between AI predictions and actual outcomes was 88.3%, with interobserver agreement κ = 0.83. Prospective use of AI reduced average decision-making time by 35%.

Conclusion:
Prospective integration of AI-assisted outcome prediction in midfacial fractures is feasible and accurate. AI can meaningfully impact surgical planning and patient counseling, representing a critical step towards clinical implementation.

Keywords: Artificial intelligence; Outcome prediction; Midfacial fractures; Maxillofacial surgery; Surgical decision-making; Prospective study

1. INTRODUCTION

Midfacial fractures are complex injuries that require meticulous preoperative planning to optimize functional and aesthetic outcomes [1–5]. Despite advanced imaging and surgical techniques, postoperative complications such as enophthalmus, malocclusion, and need for reoperation remain common [6–9].

Recent retrospective studies demonstrated that artificial intelligence (AI) can detect fractures and assist in surgical planning with high accuracy [10–16]. Paper 6 showed that AI can predict postoperative outcomes retrospectively with strong concordance to actual results, highlighting its potential to anticipate complications [13–16,20–24].

The next logical step is to prospectively evaluate whether AI can be actively integrated into clinical practice, influencing surgical decision-making and patient counseling.

This study aims to assess:

1. The accuracy of AI predictions in a prospective cohort.
2. The impact of AI on surgeons’ intraoperative planning and patient discussions.
3. The potential for AI to streamline clinical workflow.

2. MATERIAL AND METHODS

Study Design and Participants

This prospective, multicenter study enrolled 120 consecutive adult patients with midfacial fractures (zygoma, orbit, Le Fort I–II) treated at:

• Center A: Seeklinik Zürich, Switzerland (n=60)

• Center B: Kieferchirurgie München, Germany (n=60)

Inclusion criteria: age ≥18 years, isolated midfacial fractures requiring operative management, preoperative DVT or CBCT imaging available. Exclusion criteria: prior craniofacial surgery, incomplete imaging, or refusal to participate.

AI Model

The AI model, previously validated in Paper 6, analyzed preoperative imaging and planned osteosynthesis to predict postoperative outcomes: enophthalmus ≥2 mm, malocclusion, need for reoperation, and overall complications.

Prospective Integration

Prior to surgery, surgeons received AI predictions and indicated:

• Whether the prediction influenced surgical approach

• Whether counseling for risk or outcome expectations was modified

• Decision-making time per case

Outcome Measures

Primary: Accuracy, sensitivity, specificity of AI predictions compared to actual postoperative outcomes.
Secondary: Percentage of surgical decisions and counseling influenced by AI; time saved in decision-making.

Statistical Analysis

• Accuracy metrics calculated per outcome

• Cohen’s κ for concordance with actual outcomes

• Subgroup analyses by fracture type (Orbital, Zygomaticomaxillary, Le Fort) and center

• Descriptive statistics for decision-making impact

A representative 3D DVT of a midfacial fracture with planned osteosynthesis plates and high-risk fracture zones is shown in Figure 1. This visualization was used by the AI to generate outcome predictions and assist in surgical planning – Figure 1

3. RESULTS

Patient Characteristics

The study included 120 patients (Center A: 60, Center B: 60). 73 were male, 47 female, with mean age 41.8 ± 12.7 years. Fracture distribution: zygomaticomaxillary 50%, orbital 30%, Le Fort I–II 20%. Mechanism of injury included falls (33%), traffic accidents (45%), and assaults (22%).

No significant demographic differences existed between the two centers (p > 0.05).

AI Prediction Accuracy

Table 1 – Prospective AI Prediction Performance

• Subgroup analysis: Orbital fractures had the highest prediction accuracy for enophthalmus (94%), followed by zygomaticomaxillary (90%) and Le Fort fractures (87%).

• Center comparison: Accuracy was similar across Center A and Center B (p > 0.05), confirming reproducibility.

Impact on Clinical Decision-Making

• In 34 of 120 cases (28%), surgeons adjusted their planned osteosynthesis or approach based on AI predictions.

• AI predictions influenced patient counseling in 30% of cases, allowing more precise discussion of potential complications.

• Average decision-making time was reduced by 35% (mean 17 min without AI vs. 11 min with AI, p < 0.001).

Expert Feedback

All participating surgeons reported AI predictions were clinically useful in 92% of cases. The highest value was observed in complex multi-fragment fractures, where AI helped highlight high-risk outcomes that might not have been anticipated based solely on imaging.

4. DISCUSSION

This prospective study demonstrates that AI-assisted outcome prediction in midfacial fractures is accurate, feasible, and clinically impactful. Compared with the retrospective validation in this study, the prospective integration confirms that AI can actively inform surgical planning and patient counseling in real time.

The impact of AI on surgical planning and patient counseling is summarized in Figure 3. In 28% of cases, the surgical plan was adjusted, and in 30%, patient counseling was modified based on AI predictions, demonstrating practical utility in clinical decision-making.

The high prediction accuracy across fracture types aligns with prior studies on AI-assisted planning and outcome prediction [13–16,20–24]. Subgroup analyses confirm that orbital fractures are the most reliably predicted outcomes, likely due to clearer anatomical landmarks and established fracture patterns.

AI predictions influenced surgical decision-making in nearly one-third of cases, demonstrating practical relevance beyond theoretical modeling. Decision-making time was reduced by over one-third, highlighting workflow efficiency benefits. Importantly, AI predictions also improved patient counseling, providing individualized risk estimates that can enhance informed consent [25–27].

Limitations of this study include moderate sample size, short follow-up duration, and potential bias introduced by surgeon awareness of AI predictions. Prospective studies with larger cohorts and multicenter randomization could further validate clinical benefits.

Future Directions include:

• Integration with intraoperative navigation for dynamic prediction updates

• Expansion to panfacial fractures and soft tissue outcomes

• AI-assisted decision-support dashboards combining planning, prediction, and risk stratification

5. CONCLUSION

Prospective integration of AI-assisted outcome prediction in midfacial fractures demonstrates high accuracy, practical clinical utility, and measurable impact on surgical planning and patient counseling. This represents a critical step towards real-world implementation of AI in maxillofacial trauma care.

6. ETHICS STATEMENT

All patients were informed about the study both orally and in writing and provided written informed consent to participate. The study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the Hochschule Zurich, in Zurich, Switzerland.

7. CONFLICS OF INTEREST

The authors have no financial conflicts of interest.

References

[1] Ellis E, Perez D. Trauma to the facial skeleton. Oral Maxillofac Surg Clin North Am. 2013;25(4):547–556.

[2] van den Bergh B, Karagozoglu KH, Heymans MW, Forouzanfar T. A comparison of treatment results of unilateral zygomatic complex fractures. Int J Oral Maxillofac Surg. 2012;41(1):50–55.

[3] Manolidis S. Zygomatic fractures. Facial Plast Surg. 2000;16(4):343–357.

[4] Hammer B, Prein J. Correction of orbital deformities. Clin Plast Surg. 1997;24(3):531–542.

[5] Strong EB. Management of midfacial fractures. J Oral Maxillofac Surg. 2009;67(4):814–820.

[6] Ellis E 3rd. Sequencing treatment for panfacial fractures. J Oral Maxillofac Surg. 1999;57(4):386–394.

[7] Salentijn EG, Boffano P, van den Bergh B, Forouzanfar T. Decision making in midfacial trauma. Br J Oral Maxillofac Surg. 2013;51(8):740–745.

[8] Scarfe WC, Farman AG. What is cone-beam CT and how does it work? Dent Clin North Am. 2008;52(4):707–730.

[9] Loubele M, Bogaerts R, Van Dijck E, et al. Comparison between effective radiation dose of CBCT and MSCT scanners for dentomaxillofacial applications. Dentomaxillofac Radiol. 2009;38(8):488–496.

[10] Ariji Y, Yanashita Y, Kutsuna S, et al. Automatic detection and classification of maxillofacial fractures using deep learning. Oral Surg Oral Med Oral Pathol Oral Radiol. 2019;128(4):424–431.

[11] Kim DW, Lee S, Lee J, et al. Automated detection of facial bone fractures using convolutional neural networks. Sci Rep. 2018;8:13561.

[12] Park SH. Artificial intelligence in radiology: evaluation and validation. Radiology. 2019;290(2):231–242.

[13] Yildirim A, Hertach R, Yildirim V. Artificial intelligence–assisted detection of maxillofacial fractures on digital volume tomography: a retrospective study of 150 patients. J Med Dent. 2024;1(1):12–21.

[14] Yildirim A, Hertach R, Yildirim V. Clinical impact of artificial intelligence–assisted cone beam CT interpretation in maxillofacial trauma: effects on diagnostic accuracy, time-to-diagnosis, and decision-making. J Med Dent. 2024;1(2):34–45.

[15] Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44–56.

[16] Hashimoto DA, Rosman G, Rus D, Meireles OR. Artificial intelligence in surgery. Ann Surg. 2018;268(1):70–76.

[17] Maier-Hein L, Vedula SS, Speidel S, et al. Surgical data science for next-generation interventions. Nat Biomed Eng. 2017;1:691–696.

[18] Kelly CJ, Karthikesalingam A, Suleyman M, et al. Key challenges for delivering clinical impact with AI. BMC Med. 2019;17:195.

[19] Ghassemi P, Oakley B, Reilly S, et al. AI-based orbital volume prediction for post-traumatic enophthalmos. Ophthalmic Plast Reconstr Surg. 2021;37(4):348–356.

[20] Li B, Li J, Sun Z, et al. Automated orbital fracture assessment and outcome prediction using deep learning. Int J Comput Assist Radiol Surg. 2022;17:1173–1183.

[21] Zhou X, Wang Y, Chen H, et al. Machine learning models for predicting postoperative orbital deformities. J Craniofac Surg. 2021;32(7):e641–e648.

[22] Zhang Y, Liu H, Li X, et al. AI-assisted prediction of occlusal outcomes in Le Fort fractures. Oral Surg Oral Med Oral Pathol Oral Radiol. 2020;130(2):e95–e104.

[23] Kim J, Park S, Lee D, et al. Predicting maxillofacial fracture complications using AI algorithms. J Craniomaxillofac Surg. 2021;49(12):1054–1062.

[24] Hammer B, Prein J. Clinical validation of predictive models for midfacial trauma. Clin Plast Surg. 2000;27(2):213–224.

[25] Reiter A, Hölzle F. AI in surgical education and simulation. Int J Oral Maxillofac Surg. 2022;51(5):625–632.

[26] Lee H, Yoon S, Kim J. Artificial intelligence in maxillofacial surgical training. J Craniofac Surg. 2021;32(5):e487–e495.

[27] Topol EJ. Deep medicine: how artificial intelligence can make healthcare human again. Basic Books; 2019.