pith. sign in

arxiv: 2605.01084 · v1 · submitted 2026-05-01 · 💻 cs.CV

Patient-Specific Optimization for Mandibular Reconstruction Planning with Enhanced Bone Union

Hamidreza Aftabi , John E. Lloyd , Amanda Ding , Benedikt Sagl , Eitan Prisman , Antony Hodgson , Sidney Fels This is my paper

Pith reviewed 2026-05-09 19:04 UTC · model grok-4.3

classification 💻 cs.CV
keywords mandibular reconstructionvascularized bone graftbone unionvirtual surgical planningBayesian optimizationpatient-specific modelingdigital twinapposition
0
0 comments X

The pith

Patient-specific digital twins and Bayesian optimization increase donor-mandible apposition by up to 29 percentage points in mandibular reconstruction planning.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents a workflow that converts a pre-operative CT into a patient-specific digital twin by registering templates and updating muscle and joint parameters from imaging. Bayesian optimization then searches six cut-plane and donor-position variables to maximize an objective based on cycle-averaged bone apposition under a safety constraint. This produces plans that raise apposition by up to 29 percentage points over standard approaches in generic defects and 26 points over actual post-operative configurations in patient cases. A reader would care because nonunion after vascularized bone grafting remains a frequent problem, and existing virtual planning tools focus on geometry rather than mechanical conditions that support healing.

Core claim

OsteoOpt++ converts pre-operative CT data into a personalized digital twin through template-to-patient registration and CT-derived muscle and temporomandibular-joint updates, then applies Bayesian optimization with an expected-improvement-plus acquisition function to search six clinically controllable variables under an apposition-driven objective, achieving cycle-averaged donor-mandible apposition gains of up to 29 percentage points against common surgical approaches in generic cases and up to 26 percentage points against surgeon-implemented day-5 configurations in patient-specific cases.

What carries the argument

OsteoOpt++, an image-to-decision planning loop that builds a patient-specific digital twin via template registration and CT-derived tissue updates, then uses Bayesian optimization to maximize an apposition objective over six cut-plane and donor-positioning variables.

If this is right

  • Surgeons obtain pre-operative, image-driven recommendations for cut-plane orientation and donor placement that are predicted to improve union conditions over configurations delivered in the operating room.
  • The method delivers measurable apposition increases in both generic defect models and real patient-specific cases with low sensitivity to eleven modeling parameters.
  • In one longitudinal validation case the predicted apposition maps show Dice overlap of 0.70 and 0.76 with year-1 bone formation observed on follow-up imaging.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same digital-twin and optimization loop could be applied to other vascularized bone graft sites where nonunion rates are also high.
  • Real-time surgical navigation systems might incorporate the apposition objective to allow intra-operative fine-tuning of cut angles or graft seating.
  • Larger multi-center cohorts could test whether the reported apposition gains correspond to fewer secondary interventions for nonunion.

Load-bearing premise

The apposition-driven objective computed from the digital twin accurately predicts improved clinical bone union, and the template-to-patient registration together with CT-derived muscle and TMJ updates produce a sufficiently faithful biomechanical model.

What would settle it

A prospective clinical comparison of one-year radiographic bone union rates between patients whose reconstructions used the optimized cut-plane and donor-position plans versus matched patients whose plans used standard geometric virtual surgical planning.

Figures

Figures reproduced from arXiv: 2605.01084 by Amanda Ding, Antony Hodgson, Benedikt Sagl, Eitan Prisman, Hamidreza Aftabi, John E. Lloyd, Sidney Fels.

Figure 1
Figure 1. Figure 1: Planning and optimization workflow. The process starts from CT-based segmentation of the mandible and donor fibula, followed by construction of the corresponding 3D surface models and definition of the tumor or resection region. For each candidate design, the virtual-planning stage generates the defect-specific resection geometry, donor-segment configuration, and fixation-plate layout. The arrowheads indic… view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the automated patient-specific workflow. The diagram summarizes the template-to-patient pipeline described in Algorithm 1. A healthy generic craniofacial model provides the anatomical prior, while patient CT provides the patient-specific skeletal geometry and muscle information used for personalization. These two streams are coupled through maxilla-based rigid initialization, coherent point dri… view at source ↗
Figure 3
Figure 3. Figure 3: Muscle segmentation and scan cross-section plane definition for PCSA estimation. TotalSegmentator Wasserthal et al. (2023) is used to segment the principal masticatory muscles from patient CT. Landmark-defined reference planes and neighboring parallel planes are then used to extract scan cross sections (SCSs) for the PCSA update. detailed muscle fiber orientation information, which cannot be obtained from … view at source ↗
Figure 4
Figure 4. Figure 4: Optimization outcomes for the generic defect cases. Rows show B, S, and RB defects; columns show optimized design parameters mapped to [−100, 100], mean donor-mandible apposition over one chewing cycle, and the corresponding apposition trajectory. The gray interval marks bolus engagement. Results are shown for baseline reconstructions, representing common surgical practice with zero design-variable offsets… view at source ↗
Figure 5
Figure 5. Figure 5: Generic reconstruction configurations. Baseline reconstructions, representing common surgical practice, and the optimized 𝐹opt and 𝐹SF reconstructions are shown for the B, S, and RB defect cases from front, top, and bottom views. Taken together, the generic-model results indicate that the proposed cost functions are well posed, that mean￾ingful improvements over common-practice baselines are recoverable in… view at source ↗
Figure 6
Figure 6. Figure 6: Optimization behavior for the 𝐹opt objective. Panels (a)–(c) show 150-iteration parallel-coordinate analyses relating normalized design parameters to the objective value for the B, S, and RB defects, respectively. Parameters are normalized to [−100, 100], and magenta highlights the optimal region. Panel (d) shows an example convergence profile for the B defect over 75 iterations across five optimization tr… view at source ↗
Figure 7
Figure 7. Figure 7: Optimization outcomes for the patient-specific cases. Rows show the patient-specific B, S, and RB cases; columns show optimized design parameters mapped to [−100, 100], mean donor-mandible apposition over one chewing cycle, and the corresponding apposition trajectory. The gray interval marks bolus engagement. Results are shown for baseline reconstructions, defined by the surgeon-implemented day-5 post-op C… view at source ↗
Figure 8
Figure 8. Figure 8: Patient-specific reconstruction configurations. Baseline reconstructions, corresponding to the surgeon-implemented configuration recovered from day-5 post-op CT, and the optimized 𝐹opt and 𝐹SF reconstructions are shown for the patient-specific B, S, and RB cases from front, top, and bottom views. while the underlying configurations differ in their worst￾case principal stresses on cortical and cancellous bo… view at source ↗
Figure 9
Figure 9. Figure 9: Sensitivity analysis of model parameters. Average effect, over five runs, of −10% and +10% perturbations in 11 modeling parameters on the 𝐹opt objective for (a) the generic B, S, and RB defect models and (b) the patient-specific models with similar defect types. G denotes the generic model, P denotes the patient-specific model, and superscripts − and + indicate −10% and +10% perturbations, respectively. Ac… view at source ↗
Figure 10
Figure 10. Figure 10: TMJ-disc sensitivity analysis. (a) T1-weighted MRI was registered to the CT image using Elastix, and the fused image was used to guide temporomandibular joint disc segmentation. (b) Donor-mandible apposition trajectories over one chewing cycle are compared between the real segmented disc and the anatomy-guided disc approximation used in the patient-specific model. The gray interval marks bolus engagement.… view at source ↗
Figure 11
Figure 11. Figure 11: Longitudinal validation of predicted bone-union apposition. The patient-specific model was constructed from day-5 post-op CT and used to predict donor-mandible apposition near the resection interfaces. The predicted apposition pattern was compared with bone formation observed on year-1 post-op CT, yielding Dice scores of 0.70 and 0.76 at the left and right resection interfaces, respectively. This study sh… view at source ↗
read the original abstract

Mandibular reconstruction with vascularized bone grafts is complicated by donor-host nonunion, and current virtual surgical planning produces a geometric plan rather than a configuration that explicitly promotes bone union. We present OsteoOpt++, an image-to-decision planning loop for patient-specific mandibular reconstruction. A pre-operative computed tomography (CT) is converted into a personalized digital twin through template-to-patient registration and CT-derived updates of the muscle and temporomandibular-joint parameters. Bayesian optimization with an expected-improvement-plus acquisition rule then searches six clinically controllable cut-plane and donor-positioning variables under an apposition-driven objective and a safety-factor-regularized variant. The workflow was evaluated on three generic defects (body, symphysis, and ramus-body) and a total of 3+1 patient-specific cases, with 3 used for optimization and 1 for validation. In the generic cases, against a common surgical approach, cycle-averaged donor-mandible apposition increased by up to 29 percentage points (329% relative); in the patient-specific cases, against the surgeon-implemented day-5 post-operative configuration, by up to 26 percentage points. A 10% sensitivity analysis over eleven modeling parameters capped the change in the apposition-driven objective at 3% for generic cases and 4% for patient-specific cases, and the longitudinal case showed Dice overlap of 0.70 and 0.76 between predicted apposition and year-1 bone formation. Clinically, this provides surgeons with a pre-operative, image-driven recommendation for cut-plane orientation and donor placement that is predicted to improve union conditions over the configurations currently delivered in the operating room. The optimization and patient-specific modeling code is open source at https://github.com/hamidreza-aftabi/OsteoOpt.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 3 minor

Summary. The manuscript introduces OsteoOpt++, a patient-specific planning framework that converts preoperative CT into a digital twin via template registration and CT-derived muscle/TMJ updates, then applies Bayesian optimization (expected-improvement-plus) over six cut-plane and donor-positioning variables to maximize a cycle-averaged apposition objective (with a safety-factor variant). It reports quantitative gains versus standard approaches on three generic defects (up to 29 pp / 329% relative) and four patient cases (up to 26 pp versus day-5 postoperative configuration), supported by a 10% sensitivity sweep over eleven parameters and one longitudinal validation case (Dice 0.70/0.76 at year 1).

Significance. If the apposition surrogate proves reliable, the work offers a reproducible, image-driven method to optimize biomechanical conditions for bone union beyond pure geometry, with open-source code as a clear strength for verification. The small sample and single-case validation, however, constrain immediate clinical translation and generalizability.

major comments (3)
  1. [Longitudinal validation and patient-specific results] The central clinical claim—that optimized apposition improves bone union—rests on a surrogate whose validity is demonstrated in only one longitudinal patient case (Dice overlap 0.70–0.76 between predicted apposition and year-1 formation). No multi-patient statistical correlation, no direct union-rate comparison between optimized and actual OR configurations, and no external validation of the CT-derived muscle/TMJ force model are reported; this single-case evidence is load-bearing for interpreting the 29 pp and 26 pp gains as clinically meaningful.
  2. [Evaluation on generic defects and patient cases] The generic-defect and patient-specific improvements (up to 29 pp and 26 pp) are presented as point estimates without statistical tests, confidence intervals, or explicit definition of the baseline 'common surgical approach' configuration; this weakens the quantitative claims in the absence of variability measures or hypothesis testing.
  3. [Sensitivity analysis] The sensitivity analysis caps objective change at 3–4% under 10% parameter perturbation, yet the manuscript does not specify how the eleven modeling parameters were selected or whether they encompass all sources of CT segmentation and registration uncertainty; without this, the robustness statement cannot be fully assessed.
minor comments (3)
  1. [Abstract] The abstract phrasing 'a total of 3+1 patient-specific cases, with 3 used for optimization and 1 for validation' is ambiguous; state the exact total and assignment clearly.
  2. [Figures] Figure legends and axis labels should explicitly indicate whether apposition values are cycle-averaged and whether error bars represent sensitivity bounds or inter-case variability.
  3. [Code availability] The open-source repository link is provided; confirm that the released code includes the exact Bayesian optimization implementation and digital-twin construction routines used for the reported results.

Simulated Author's Rebuttal

3 responses · 2 unresolved

We thank the referee for their constructive and detailed review. The comments have helped us clarify the scope of our claims and strengthen the presentation of limitations. We respond point-by-point below and indicate the revisions made.

read point-by-point responses

  1. Referee: [Longitudinal validation and patient-specific results] The central clinical claim—that optimized apposition improves bone union—rests on a surrogate whose validity is demonstrated in only one longitudinal patient case (Dice overlap 0.70–0.76 between predicted apposition and year-1 formation). No multi-patient statistical correlation, no direct union-rate comparison between optimized and actual OR configurations, and no external validation of the CT-derived muscle/TMJ force model are reported; this single-case evidence is load-bearing for interpreting the 29 pp and 26 pp gains as clinically meaningful.

    Authors: We agree that the surrogate validation rests on a single longitudinal case, which is a genuine limitation given the scarcity of complete pre- and post-operative imaging series with long-term follow-up in mandibular reconstruction. We have added a dedicated Limitations section that explicitly states the single-case nature of the Dice validation (0.70/0.76), clarifies that the 29 pp and 26 pp figures represent improvements in the apposition surrogate rather than measured union rates, and calls for future multi-center studies to obtain statistical correlations. The muscle/TMJ parameters follow established CT-derived methods from prior literature (now cited); no separate external validation cohort was available for this study, and this is now noted as a limitation. revision: partial

  2. Referee: [Evaluation on generic defects and patient cases] The generic-defect and patient-specific improvements (up to 29 pp and 26 pp) are presented as point estimates without statistical tests, confidence intervals, or explicit definition of the baseline 'common surgical approach' configuration; this weakens the quantitative claims in the absence of variability measures or hypothesis testing.

    Authors: We have revised the Methods and Results sections to explicitly define the baseline 'common surgical approach' as the standard virtual surgical planning configuration that aligns osteotomy planes to anatomical landmarks without optimization. With only three generic defects and four patient cases and deterministic optimization per case, formal hypothesis testing or confidence intervals would be underpowered; we therefore present the gains as observed point estimates and have added a sentence noting the absence of variability measures across repeated configurations. revision: partial

  3. Referee: [Sensitivity analysis] The sensitivity analysis caps objective change at 3–4% under 10% parameter perturbation, yet the manuscript does not specify how the eleven modeling parameters were selected or whether they encompass all sources of CT segmentation and registration uncertainty; without this, the robustness statement cannot be fully assessed.

    Authors: We have expanded the Methods section to describe the selection of the eleven parameters: they were chosen from the set of biomechanical and registration variables shown to be most influential in our prior mandibular modeling work and in the broader literature on CT-based force estimation. We now state that the 10 % perturbation provides an initial robustness check rather than exhaustive coverage of all uncertainties (e.g., inter-observer CT segmentation variability). A fuller uncertainty quantification is identified as future work. revision: yes

standing simulated objections not resolved
  • Multi-patient statistical correlation or direct union-rate comparison between optimized and actual OR configurations, as these require additional longitudinal clinical datasets that are not available in the current study.
  • External validation of the CT-derived muscle/TMJ force model on an independent cohort.

Circularity Check

0 steps flagged

No circularity: optimization outputs are empirical deltas in the modeled objective, not tautological redefinitions

full rationale

The paper constructs a digital twin from CT via registration and parameter updates, then applies Bayesian optimization over six cut-plane and positioning variables to maximize an apposition-driven objective (with a safety-regularized variant). Reported improvements are measured post-optimization as differences in this objective versus baseline configurations (common surgical approach or day-5 post-op). No equation defines the objective in terms of the reported deltas or vice versa; the deltas are search outcomes, not inputs. The single-case Dice validation (0.70/0.76) and 10%-parameter sensitivity sweep are independent checks outside the optimization loop. No self-citations, uniqueness theorems, or ansatzes are invoked to force the result. The chain is therefore self-contained against its external benchmarks.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 1 invented entities

The central claim depends on the unproven premise that increased modeled apposition will translate to better clinical union rates and that the digital twin faithfully captures the relevant anatomy and mechanics for optimization purposes.

free parameters (2)
  • six clinically controllable cut-plane and donor-positioning variables
    Decision variables searched by Bayesian optimization; their values are chosen to maximize the objective rather than fitted to outcome data.
  • eleven modeling parameters
    Subjected to 10% sensitivity analysis; changes affect the apposition objective by at most 4%.
axioms (2)
  • domain assumption Donor-mandible apposition promotes bone union
    This premise directly defines the optimization objective and the claimed clinical benefit.
  • domain assumption Template-to-patient registration and CT-derived muscle/TMJ updates produce an accurate biomechanical digital twin
    Required for the patient-specific modeling step to be valid.
invented entities (1)
  • personalized digital twin no independent evidence
    purpose: To represent patient-specific anatomy, muscles, and temporomandibular joint for optimization
    Constructed via registration and CT updates; no independent falsifiable evidence provided beyond the sensitivity check.

pith-pipeline@v0.9.0 · 5645 in / 1544 out tokens · 65675 ms · 2026-05-09T19:04:13.492970+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

194 extracted references · 194 canonical work pages · 2 internal anchors

  1. [1]

    International journal of oral and maxillofacial surgery , volume=

    Biomechanics of mandibular reconstruction: a review , author=. International journal of oral and maxillofacial surgery , volume=. 2010 , publisher=

    work page 2010

  2. [2]

    Medical Image Computing and Computer Assisted Intervention--MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13--17, 2019, Proceedings, Part V 22 , pages=

    Variational shape completion for virtual planning of jaw reconstructive surgery , author=. Medical Image Computing and Computer Assisted Intervention--MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13--17, 2019, Proceedings, Part V 22 , pages=. 2019 , organization=

    work page 2019

  3. [3]

    Machine Learnt Treatment: Machine Learning and Registration Techniques for Digitally Planned Jaw Reconstructive Surgery , author=

    work page

  4. [4]

    Computers in Biology and Medicine , volume=

    Computational models and their applications in biomechanical analysis of mandibular reconstruction surgery , author=. Computers in Biology and Medicine , volume=. 2024 , publisher=

    work page 2024

  5. [5]

    Computer Methods and Programs in Biomedicine , volume=

    To what extent can mastication functionality be restored following mandibular reconstruction surgery? A computer modeling approach , author=. Computer Methods and Programs in Biomedicine , volume=. 2024 , publisher=

    work page 2024

  6. [6]

    2024 , note=

    Restoration of Mastication Functionality Post-Mandibular Reconstruction: Insights from Dynamic Computer Modeling , author=. 2024 , note=

    work page 2024

  7. [7]

    2024 , note=

    Predicting TMJ Disc Response to Jaw Reconstruction Surgery: A Computer Modeling Approach , author=. 2024 , note=

    work page 2024

  8. [8]

    ISBI , pages=

    Optimizing bone cuts enhances predicted bone union propensity in mandibular body reconstruction , author=. ISBI , pages=

    work page

  9. [9]

    MICCAI , pages=

    OsteoOpt: A Bayesian optimization framework for enhancing bone union likelihood in mandibular reconstruction surgery , author=. MICCAI , pages=

    work page

  10. [10]

    2026 , note=

    Advancing Mandibular Reconstruction Surgery Through Computational Modeling: Bridging Surgery and Functional Recovery , author=. 2026 , note=

    work page 2026

  11. [11]

    2026 , eprint=

    OsteoFlow: Lyapunov-Guided Flow Distillation for Predicting Bone Remodeling after Mandibular Reconstruction , author=. 2026 , eprint=

    work page 2026

  12. [12]

    2023 , note=

    A Dynamic Finite Element Model of the Reconstructed Mandible With Body and Ramus Defects for Stress Analysis During Chewing , author=. 2023 , note=

    work page 2023

  13. [13]

    Journal of Cranio-Maxillofacial Surgery , volume=

    Four-dimensional computed tomography evaluation of jaw movement following mandibular reconstruction: a pilot study , author=. Journal of Cranio-Maxillofacial Surgery , volume=. 2016 , publisher=

    work page 2016

  14. [14]

    Biomechanics of Living Organs , editor=

    FRANK: A Hybrid 3D Biomechanical Model of the Head and Neck , author=. Biomechanics of Living Organs , editor=. 2017 , publisher=

    work page 2017

  15. [15]

    Medical engineering & physics , volume=

    Multibody dynamic simulation of knee contact mechanics , author=. Medical engineering & physics , volume=. 2004 , publisher=

    work page 2004

  16. [16]

    Journal of biomechanics , volume=

    Articular contact in a three-dimensional model of the knee , author=. Journal of biomechanics , volume=. 1991 , publisher=

    work page 1991

  17. [17]

    Acta Otorhinolaryngologica Italica , volume=

    Evaluation of three-dimensional mandibular movements after reconstruction with free fibula flap , author=. Acta Otorhinolaryngologica Italica , volume=. 2015 , publisher=

    work page 2015

  18. [18]

    Frontiers in Physiology , volume=

    Masticatory adaptation to occlusal changes , author=. Frontiers in Physiology , volume=. 2020 , publisher=

    work page 2020

  19. [19]

    Radiation oncology journal , volume=

    Late side effects of radiation treatment for head and neck cancer , author=. Radiation oncology journal , volume=. 2020 , publisher=

    work page 2020

  20. [20]

    British Journal of Oral and Maxillofacial Surgery , volume=

    Mandibular reconstruction with vascularised bone flaps: a systematic review over 25 years , author=. British Journal of Oral and Maxillofacial Surgery , volume=. 2017 , publisher=

    work page 2017

  21. [21]

    Controllable flow matching for 3D contrast-enhanced brain

    Chang, Heng and Shang, Yu and Wang, Haifeng and Liang, Yuxia and Wang, Haoyu and Wang, Fan and Niu, Chen and Lian, Chunfeng , booktitle=. Controllable flow matching for 3D contrast-enhanced brain

    work page

  22. [22]

    Clinical Oral Investigations , pages=

    Changes in condylar position and morphology after mandibular reconstruction by vascularized fibular free flap with condyle preservation , author=. Clinical Oral Investigations , pages=. 2023 , publisher=

    work page 2023

  23. [23]

    MICCAI , pages=

    Anatomic-constrained medical image synthesis via physiological density sampling , author=. MICCAI , pages=

    work page

  24. [24]

    , author=

    Meshlab: an open-source mesh processing tool. , author=. Eurographics Italian chapter conference , volume=. 2008 , organization=

    work page 2008

  25. [25]

    Wound Repair and Regeneration , volume=

    Biomechanical behavior of scar tissue and uninjured skin in a porcine model , author=. Wound Repair and Regeneration , volume=. 2009 , publisher=

    work page 2009

  26. [26]

    The Journal of Prosthetic Dentistry , volume=

    Physical problems in obtaining records of the maxillofacial patient , author=. The Journal of Prosthetic Dentistry , volume=. 1975 , publisher=

    work page 1975

  27. [27]

    Head & Neck: Journal for the Sciences and Specialties of the Head and Neck , volume=

    A comparison of masticatory function in patients with or without reconstruction of the mandible , author=. Head & Neck: Journal for the Sciences and Specialties of the Head and Neck , volume=. 1997 , publisher=

    work page 1997

  28. [28]

    The Journal of prosthetic dentistry , volume=

    Modeling of jaw biomechanics in the reconstructed mandibulectomy patient , author=. The Journal of prosthetic dentistry , volume=. 1999 , publisher=

    work page 1999

  29. [29]

    Qualitative assessment of newly formed bone after distraction osteogenesis of mandible in patients with facial asymmetry using 3-dimensional computed tomography , author=. J. Oral Biol. Craniofac. Res. , volume=

    work page

  30. [30]

    Clinical Oral Implants Research , volume=

    Electromyographic evaluation of implant-supported prostheses in hemimandibulectomy-reconstructed patients , author=. Clinical Oral Implants Research , volume=. 2007 , publisher=

    work page 2007

  31. [31]

    Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology , volume=

    Adaptation of jaw closing muscles after surgical mandibular advancement procedures in different vertical craniofacial types: a magnetic resonance imaging study , author=. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology , volume=. 2007 , publisher=

    work page 2007

  32. [32]

    Seminars in surgical oncology , volume=

    Mandible reconstruction with microvascular surgery , author=. Seminars in surgical oncology , volume=. 2000 , organization=

    work page 2000

  33. [33]

    Journal of mechanics in medicine and biology , volume=

    A review of finite element applications in oral and maxillofacial biomechanics , author=. Journal of mechanics in medicine and biology , volume=. 2018 , publisher=

    work page 2018

  34. [34]

    2005 , publisher=

    Gray's anatomy for students , author=. 2005 , publisher=

    work page 2005

  35. [35]

    Oral Surgery, Oral Medicine, Oral Pathology , volume=

    Studies on masticatory functions in patients with surgical mandibular reconstruction , author=. Oral Surgery, Oral Medicine, Oral Pathology , volume=. 1972 , publisher=

    work page 1972

  36. [36]

    Journal of biomechanics , volume=

    Considerations for reporting finite element analysis studies in biomechanics , author=. Journal of biomechanics , volume=. 2012 , publisher=

    work page 2012

  37. [37]

    Applied Sciences , volume=

    Implant Model Generation Method for Mandibular Defect Based on Improved 3D Unet , author=. Applied Sciences , volume=. 2023 , publisher=

    work page 2023

  38. [38]

    Journal of the Mechanical Behavior of Biomedical Materials , volume=

    Optimal placement of fixation system for scaffold-based mandibular reconstruction , author=. Journal of the Mechanical Behavior of Biomedical Materials , volume=. 2022 , publisher=

    work page 2022

  39. [39]

    Journal of biomechanics , volume=

    Prediction of mandibular bone remodelling induced by fixed partial dentures , author=. Journal of biomechanics , volume=. 2010 , publisher=

    work page 2010

  40. [40]

    Materials Science and Engineering: C , volume=

    Mechanobiologically optimization of a 3D titanium-mesh implant for mandibular large defect: A simulated study , author=. Materials Science and Engineering: C , volume=. 2019 , publisher=

    work page 2019

  41. [41]

    Numerische Mathematik , volume=

    Singular value decomposition and least squares solutions , author=. Numerische Mathematik , volume=. 1970 , doi=

    work page 1970

  42. [42]

    British Journal of Oral and Maxillofacial Surgery , volume=

    Surgical planning and microvascular reconstruction of the mandible with a fibular flap using computer-aided design, rapid prototype modelling, and precontoured titanium reconstruction plates: a prospective study , author=. British Journal of Oral and Maxillofacial Surgery , volume=. 2015 , publisher=

    work page 2015

  43. [43]

    International journal of oral and maxillofacial surgery , volume=

    Mandibular reconstruction in adults: a review , author=. International journal of oral and maxillofacial surgery , volume=. 2008 , publisher=

    work page 2008

  44. [44]

    Advances in Neural Information Processing Systems , volume=

    Generative Adversarial Nets , author=. Advances in Neural Information Processing Systems , volume=

    work page

  45. [45]

    Irbm , volume=

    Quantitative ultrasound assessment of cortical bone properties beyond bone mineral density , author=. Irbm , volume=. 2019 , publisher=

    work page 2019

  46. [46]

    The European Journal of Orthodontics , volume=

    The adaptive response of jaw muscles to varying functional demands , author=. The European Journal of Orthodontics , volume=. 2009 , publisher=

    work page 2009

  47. [47]

    2021 , publisher=

    Development of an Image-Guided Surgical Workflow and Tracked Surgical Tools for Mandibular Reconstruction Surgery , author=. 2021 , publisher=

    work page 2021

  48. [48]

    Journal of Biomechanics , volume=

    EMG-assisted forward dynamics simulation of subject-specific mandible musculoskeletal system , author=. Journal of Biomechanics , volume=. 2022 , publisher=

    work page 2022

  49. [49]

    Medical Image Analysis , volume=

    Automated planning of mandible reconstruction with fibula free flap based on shape completion and morphometric descriptors , author=. Medical Image Analysis , volume=. 2025 , publisher=

    work page 2025

  50. [50]

    MAISI: Medical

    Guo, Pengfei and Zhao, Can and Yang, Dong and Xu, Ziyue and Nath, Vishwesh and Tang, Yucheng and Simon, Benjamin and Belue, Mason and Harmon, Stephanie and Turkbey, Baris and others , booktitle=. MAISI: Medical

    work page

  51. [51]

    The laryngoscope , volume=

    Computer-assisted design and rapid prototype modeling in microvascular mandible reconstruction , author=. The laryngoscope , volume=. 2013 , publisher=

    work page 2013

  52. [52]

    ACM Trans

    TetGen, a Delaunay-based quality tetrahedral mesh generator , author=. ACM Trans. Math. Softw , volume=

    work page

  53. [53]

    Journal of Biomechanics , volume=

    A dynamic model of jaw and hyoid biomechanics during chewing , author=. Journal of Biomechanics , volume=. 2008 , publisher=

    work page 2008

  54. [54]

    The Journal of Prosthetic Dentistry , volume=

    A comparison of simulated jaw dynamics in models of segmental mandibular resection versus resection with alloplastic reconstruction , author=. The Journal of Prosthetic Dentistry , volume=. 2010 , publisher=

    work page 2010

  55. [55]

    Journal of medical and dental sciences , volume=

    Electromyographic activity of masticatory muscles and mandibular movement during function in marginal mandibulectomy patients , author=. Journal of medical and dental sciences , volume=. 2003 , publisher=

    work page 2003

  56. [56]

    Journal of biomechanics , volume=

    Modeling the biomechanics of the mandible: a three-dimensional finite element study , author=. Journal of biomechanics , volume=. 1992 , publisher=

    work page 1992

  57. [57]

    and Xu, Daguang , booktitle=

    Hatamizadeh, Ali and Nath, Vishwesh and Tang, Yucheng and Yang, Dong and Roth, Holger R. and Xu, Daguang , booktitle=. Swin

    work page

  58. [58]

    Proceedings of the Royal Society of London

    The mechanics of active muscle , author=. Proceedings of the Royal Society of London. Series B-Biological Sciences , volume=. 1953 , publisher=

    work page 1953

  59. [59]

    Maxillofacial Prosthetics , volume=

    Evaluation of stomatognathic functions in mandibulectomy patients with six degree of freedom jaw movement tracking device , author=. Maxillofacial Prosthetics , volume=

    work page

  60. [60]

    Scientific Reports , volume=

    CBCT-to-CT synthesis using a hybrid U-Net diffusion model based on transformers and information bottleneck theory , author=. Scientific Reports , volume=. 2025 , doi=

    work page 2025

  61. [61]

    Denoising diffusion probabilistic models , author=. Adv. Neural Inf. Process. Syst. , volume=

    work page

  62. [62]

    International journal of oral and maxillofacial surgery , volume=

    Rehabilitation after mandibular reconstruction with fibula free-flap: clinical outcome and quality of life assessment , author=. International journal of oral and maxillofacial surgery , volume=. 2008 , publisher=

    work page 2008

  63. [63]

    CVPR , pages=

    Image-to-image translation with conditional adversarial networks , author=. CVPR , pages=

    work page

  64. [64]

    Journal of Materials Science: Materials in Medicine , volume=

    3D printed models in mandibular reconstruction with bony free flaps , author=. Journal of Materials Science: Materials in Medicine , volume=. 2018 , publisher=

    work page 2018

  65. [65]

    ISM , pages=

    ResUNet++: An advanced architecture for medical image segmentation , author=. ISM , pages=

    work page

  66. [66]

    Jiang, Caiwen and Xing, Xiaodan and Ou, Zaixin and Liu, Mianxin and Simon, Walsh and Yang, Guang and Shen, Dinggang , booktitle=

    work page

  67. [67]

    Maxillofacial Plastic and Reconstructive Surgery , volume=

    Condyle dislocation following mandibular reconstruction using a fibula free flap: complication cases , author=. Maxillofacial Plastic and Reconstructive Surgery , volume=. 2019 , publisher=

    work page 2019

  68. [68]

    Medical image analysis , volume=

    Deep learning based synthesis of MRI, CT and PET: Review and analysis , author=. Medical image analysis , volume=. 2024 , publisher=

    work page 2024

  69. [69]

    Medical Image Analysis , pages=

    Diffusion models in medical imaging: A comprehensive survey , author=. Medical Image Analysis , pages=. 2023 , publisher=

    work page 2023

  70. [70]

    WACV , pages=

    Adaptive latent diffusion model for 3D medical image-to-image translation: Multi-modal magnetic resonance imaging study , author=. WACV , pages=

    work page

  71. [71]

    Journal of Cranio-Maxillofacial Surgery , volume=

    Adaquate fixation of plates for stability during mandibular reconstruction , author=. Journal of Cranio-Maxillofacial Surgery , volume=. 2006 , publisher=

    work page 2006

  72. [72]

    Auto-Encoding Variational Bayes

    Auto-Encoding Variational Bayes , author=. arXiv preprint arXiv:1312.6114 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  73. [73]

    IEEE transactions on medical imaging , volume=

    Elastix: a toolbox for intensity-based medical image registration , author=. IEEE transactions on medical imaging , volume=. 2009 , publisher=

    work page 2009

  74. [74]

    Patient-specific implants: A retrospective analysis of 89 patients , author=

    Osseous union after mandible reconstruction with fibula free flap using manually bent plates vs. Patient-specific implants: A retrospective analysis of 89 patients , author=. Current Oncology , volume=. 2022 , publisher=

    work page 2022

  75. [75]

    MICCAI , pages=

    Anatomically-controllable medical image generation with segmentation-guided diffusion models , author=. MICCAI , pages=

    work page

  76. [76]

    journal of the mechanical behavior of biomedical materials , volume=

    Semi-automated digital workflow to design and evaluate patient-specific mandibular reconstruction implants , author=. journal of the mechanical behavior of biomedical materials , volume=. 2022 , publisher=

    work page 2022

  77. [77]

    journal of the mechanical behavior of biomedical materials , volume=

    Topology optimization of a mandibular reconstruction plate and biomechanical validation , author=. journal of the mechanical behavior of biomedical materials , volume=. 2021 , publisher=

    work page 2021

  78. [78]

    The Journal of Prosthetic Dentistry , volume=

    Effect of bilateral asymmetric tooth clenching on load distribution at the mandibular condyles , author=. The Journal of Prosthetic Dentistry , volume=. 1990 , publisher=

    work page 1990

  79. [79]

    Journal of oral rehabilitation , volume=

    Influence of surgical orthodontic treatment on masticatory function in skeletal C lass III patients , author=. Journal of oral rehabilitation , volume=. 2015 , publisher=

    work page 2015

  80. [80]

    Journal of maxillofacial and oral surgery , volume=

    Mandibular reconstruction: overview , author=. Journal of maxillofacial and oral surgery , volume=. 2016 , publisher=

    work page 2016

Showing first 80 references.