背景：颌骨缺损的精准重建与功能修复是颌面整形外科领域所面临的艰巨挑战。传统游离骨瓣移植术存在截骨重建精度不足、术后功能恢复有限等问题，而恶性肿瘤术后放疗引发的骨代谢障碍进一步制约种植修复的临床应用。针对上述问题，本研究以“精准化”与“功能化”为主线，从数字化外科技术创新和放疗后种植体骨结合规律出发展开研究，旨在从解剖重建到功能恢复进行多维度的探索，为提升患者生存质量提供理论依据与技术支持。

方法：本研究包括以下三个部分:（1）改良截骨就位一体化导板系统（Modified Osteotomy and Positioning Integrative Template System, MOPITS）的临床验证：纳入70例颌骨缺损患者，采用MOPITS系统完成颌骨缺损重建手术。通过CT重建模型评估截骨精度与下颌骨融合误差。随访3个月至5年，采用华盛顿大学生活质量量表（University of Washington Quality of Life Questionnaire, UW-QOL）、功能性经口摄食量表（Functional Oral Intake Scale，FOIS）及美国足踝外科协会（American Orthopaedic Foot & Ankle Society，AOFAS）踝－后足评分量表进行评估。（2）人机协同增强现实导航系统（Human-Machine Collaborative Augmented Reality-Based Mandibular Reconstruction Surgical Navigation System，HMARS）的开发与实验验证：基于9组下颌骨-腓骨标本及2组尸头-尸腿标本，构建HMARS系统，融合机器人辅助截骨（Robot-Assisted Guided Osteotomy，RAO）与增强现实引导重建（Augmented Reality Guided Reconstruction, ARR）。通过Super4PCS算法实现术前规划与术中场景的动态配准，并设计旋转卡尺补偿法（Rotating Calipers-based Compensation，RCC）优化AR融合误差。定量评估目标配准误差（Target Registration Error，TRE）、截骨精度及重建融合误差。（3）放疗后种植体骨结合规律的动物模型研究：建立比格犬股骨种植模型，模拟临床腓骨瓣移植同期种植后放疗场景。6只比格犬分为未放疗组、放疗后2周组及4周组，每组植入8颗快速骨愈合种植体，实验组于术后6周接受8.5×3Gy的放疗。通过影像学、种植体稳定性测试及骨组织形态计量学评估放疗对骨结合的影响。

结果：（1）MOPITS系统可实现高精度截骨，腓骨截骨长度误差1.05±1.03 mm、截骨面角度误差1.68±1.40°；下颌骨整体融合误差为1.32±0.49 mm，关键标志点偏差<2 mm。UW-QOL评分各单项均＞75分，其中咀嚼、外貌、语言对患者术后影响最大。FOIS评分44.1%的患者恢复正常经口进食无任何限制，腓骨供区AOFAS评分达91.37±11.13分，并发症发生率可控。（2）HMARS系统的AR融合误差经RCC优化后从2.23±0.54 mm降至1.08±0.43 mm。腓骨截骨长度误差为1.65±1.66 mm，下颌骨重建绝对误差1.50±0.48 mm，关键解剖标志点误差<5 mm，验证了人机协同导航的临床潜力。（3）放疗后种植体骨结合的动态规律显示，放疗后2周组骨小梁面积百分比、骨小梁数量较未放疗组下降28.6%与34.1%，骨小梁分离度增加47.8%；放疗后4周组骨量恢复至未放疗组的86.7%。双皮质植入策略显著提升髓质骨种植体骨结合率。

结论: 本研究从解剖精度优化（MOPITS）、操作智能化（HMARS）到咀嚼功能重建（放疗后种植体骨结合）逐层推进，探索了颌骨缺损精准重建与功能修复。MOPITS系统通过标准化设计流程，将截骨误差控制在1.5 mm内，为复杂缺损修复提供高效方案；HMARS系统提出人机协同导航模式，AR动态配准误差≤1.1 mm，突破传统导板的静态局限；同期植入快速骨愈合种植体策略初步揭示了放疗后骨结合的“抑制-代偿”规律，为放疗患者的功能性修复提供理论支持。未来将进一步通过多中心研究验证技术普适性，并探索人工智能驱动的个性化修复路径，推动颌面缺损治疗迈向精准功能重建的新阶段。

Background: The precise reconstruction and functional rehabilitation of mandibular defects remain formidable challenges in maxillofacial plastic surgery. Traditional free bone flap transplantation suffers from insufficient osteotomy accuracy and limited postoperative functional recovery, while post-radiotherapy bone metabolism disorders following malignant tumor surgery further restrict the clinical application of implant rehabilitation. To address these issues, this study focuses on "precision" and "functionalization," exploring innovations in digital surgical technology and the osseointegration patterns of implants post-radiotherapy. The research aims to advance multidimensional progress from anatomical reconstruction to functional recovery, providing theoretical foundations and technical support to improve patients’ quality of life.

Methods: This study comprises three components: (1) Clinical validation of the Modified Osteotomy and Positioning Integrative Template System (MOPITS): Seventy patients with mandibular defects underwent reconstruction using MOPITS. CT-reconstructed models were used to evaluate osteotomy accuracy and mandibular fusion errors. Follow-up (3 months to 5 years) included assessments using the University of Washington Quality of Life Questionnaire (UW-QOL), Functional Oral Intake Scale (FOIS), and American Orthopaedic Foot & Ankle Society (AOFAS) ankle-hindfoot scale. (2) Development and experimental validation of the Human-Machine Collaborative Augmented Reality-Based Mandibular Reconstruction Surgical Navigation System (HMARS): Utilizing 9 mandible-fibula specimens and 2 cadaver head-leg specimens, the HMARS system integrated robot-assisted guided osteotomy (RAO) and augmented reality-guided reconstruction (ARR). The Super4PCS algorithm enabled dynamic registration of preoperative plans and intraoperative scenarios, while the Rotating Calipers-based Compensation (RCC) method optimized AR fusion errors. Target Registration Error (TRE), osteotomy accuracy, and reconstruction fusion errors were quantitatively assessed. (3) Animal model study on osseointegration patterns of implants post-radiotherapy: A Beagle dog femoral implant model simulated clinical scenarios of fibula flap transplantation with concurrent implant placement followed by radiotherapy. Six Beagles were divided into non-radiotherapy, 2-week post-radiotherapy, and 4-week post-radiotherapy groups, each receiving 8 rapid osseointegration implants. The experimental group underwent 8.5×3 Gy radiotherapy 6 weeks postoperatively. Radiographic, implant stability, and histomorphometric analyses evaluated radiotherapy’s impact on osseointegration.

Results: (1) MOPITS achieved high-precision osteotomy, with fibular osteotomy length error of 1.05±1.03 mm, angle error of 1.68±1.40°, and mandibular fusion error of 1.32±0.49 mm (key landmark deviations <2 mm). UW-QOL scores exceeded 75 points across domains, with chewing, appearance, and speech most impacting postoperative outcomes. FOIS scores showed 44.1% of patients regained unrestricted oral intake, and AOFAS scores for fibula donor sites reached 91.37±11.13, with manageable complications. (2) HMARS reduced AR fusion errors from 2.23±0.54 mm to 1.08±0.43 mm after RCC optimization. Fibular osteotomy length error was 1.65±1.66 mm, mandibular reconstruction absolute error 1.50±0.48 mm, and key landmark deviations <5 mm, demonstrating clinical potential. (3) Post-radiotherapy osseointegration revealed trabecular bone area percentage and trabecular number decreased by 28.6% and 34.1% (2-week group vs. non-radiotherapy), while trabecular separation increased by 47.8%. Bone volume recovered to 86.7% of non-radiotherapy levels by 4 weeks. Bicortical implantation significantly enhanced medullary bone osseointegration.

Conclusion: This study advances precision reconstruction and functional rehabilitation of mandibular defects through anatomical accuracy optimization (MOPITS), operational intelligence (HMARS), and masticatory function restoration (post-radiotherapy osseointegration). MOPITS standardizes design workflows, controlling osteotomy errors within 1.5 mm for complex defect repair. HMARS introduces human-machine collaborative navigation, achieving AR dynamic registration errors ≤1.1 mm, overcoming static limitations of traditional guides. Concurrent rapid osseointegration implant strategies reveal a "suppression-compensation" pattern post-radiotherapy, supporting functional rehabilitation in irradiated patients. Future multicenter studies will validate universal applicability and explore AI-driven personalized reconstruction pathways, advancing maxillofacial defect treatment into a new era of precision functional restoration.