Next Article in Journal
Distal Nerve Transfers in High Peroneal Nerve Lesions: An Anatomical Feasibility Study
Next Article in Special Issue
A New Approach to Virtual Occlusion in Orthognathic Surgery Planning Using Mixed Reality—A Technical Note and Review of the Literature
Previous Article in Journal
CT Perfusion as a Predictor of the Final Infarct Volume in Patients with Tandem Occlusion
Previous Article in Special Issue
Application of Augmented Reality to Maxillary Resections: A Three-Dimensional Approach to Maxillofacial Oncologic Surgery
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
JPM
Volume 13
Issue 2
10.3390/jpm13020343
jpm-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Semi-Automatic Approach for Holistic 3D Assessment of Temporomandibular Joint Changes

by
Michael Boelstoft Holte
1,2,*,
Henrik Sæderup
1 and
Else Marie Pinholt
1,2
1
3D Lab Denmark, Department of Oral and Maxillofacial Surgery, University Hospital of Southern Denmark, Finsensgade 35, 6700 Esbjerg, Denmark
2
Department of Regional Health Research, Faculty of Health Sciences, University of Southern Denmark, Finsensgade 35, 6700 Esbjerg, Denmark
*
Author to whom correspondence should be addressed.
J. Pers. Med. 2023, 13(2), 343; https://doi.org/10.3390/jpm13020343
Submission received: 10 January 2023 / Revised: 10 February 2023 / Accepted: 13 February 2023 / Published: 16 February 2023
(This article belongs to the Special Issue Personalized Treatment of Maxillofacial Deformities and Reconstruction: Emerging Technologies, Methods and Applications)
Browse Figures
Review Reports Versions Notes

Abstract

:
The literature lacks a reliable holistic approach for the three-dimensional (3D) assessment of the temporomandibular joint (TMJ) including all three adaptive processes, which are believed to contribute to the position of the mandible: (1) adaptive condylar changes, (2) glenoid fossa changes, and (3) condylar positional changes within the fossa. Hence, the purpose of the present study was to propose and assess the reliability of a semi-automatic approach for a 3D assessment of the TMJ from cone-beam computed tomography (CBCT) following orthognathic surgery. The TMJs were 3D reconstructed from a pair of superimposed pre- and postoperative (two years) CBCT scans, and spatially divided into sub-regions. The changes in the TMJ were calculated and quantified by morphovolumetrical measurements. To evaluate the reliability, intra-class correlation coefficients (ICC) were calculated at a 95% confidence interval on the measurements of two observers. The approach was deemed reliable if the ICC was good (>0.60). Pre- and postoperative CBCT scans of ten subjects (nine female; one male; mean age 25.6 years) with class II malocclusion and maxillomandibular retrognathia, who underwent bimaxillary surgery, were assessed. The inter-observer reliability of the measurements on the sample of the twenty TMJs was good to excellent, ICC range (0.71–1.00). The range of the mean absolute difference of the repeated inter-observer condylar volumetric and distance measurements, glenoid fossa surface distance measurements, and change in minimum joint space distance measurements were (1.68% (1.58)–5.01% (3.85)), (0.09 mm (0.12)–0.25 mm (0.46)), (0.05 mm (0.05)–0.08 mm (0.06)) and (0.12 mm (0.09)–0.19 mm (0.18)), respectively. The proposed semi-automatic approach demonstrated good to excellent reliability for the holistic 3D assessment of the TMJ including all three adaptive processes.

1. Introduction

Three-dimensional (3D) assessment of postoperative condylar resorption and remodeling following orthognathic surgery using superimposed cone beam computed tomography (CBCT) imaging has been studied and published numerous times [1,2,3]. In contrast, only a single recent study has been conducted assessing 3D glenoid fossa changes following orthognathic surgery [4]. The study found significant glenoid fossa changes, and subjects with postoperative condylar resorption showed a significantly higher degree of morphological change in the anterior glenoid fossa than subjects without [4].
Studies suggest 3D superimposition methods to assess glenoid fossa changes in growing subjects [5,6], and to assess the joint space [7,8]. However, none of the applied methods have been validated. Superimposition of CBCT enables automation of 3D dentomaxillofacial assessment and visualization without the need for cephalometric landmark re-identification [1,9,10,11]. Surface-based registration (SBR) comprises the approximation of two surfaces by selection and alignment using the iterative closest point algorithm [12]. Voxel-based registration (VBR) [13,14] uses the grayscale of the voxels to align two 3D volumetric images to the best superimposition [15]. It is today’s method of choice [12,16,17,18] and is considered more reliable than landmark-based cephalometric analysis [16]. Alignment by VBR, validated with the anterior cranial base as a reference, shows high accuracy [19], and is considered to be more consistent than SBR [12,19]. However, a recent study by Holte et al. [20], validated and compared VBR and SBR on the mandibular rami for long-term 3D evaluation of condylar remodeling following orthognathic surgery. In this particular case, SBR was found to be more accurate and more reliable than VBR, possibly due to the influence of mandibular ramal remodeling [20].
Three adaptive processes in the temporomandibular joint (TMJ) are believed to contribute to the position of the mandible: (1) adaptive condylar changes, (2) glenoid fossa changes, and (3) condylar positional changes within the fossa [21]. Adaptive morphovolumetrical and positional changes of the TMJ are thought to be induced during mastication or mechanical loading, e.g., by orthodontic treatment, orthognathic surgery, trauma, or general disease such as rheumatoid arthritis, where the condyles may be displaced in the glenoid fossa [22].
In order to study and understand the postoperative stability and functionality following orthognathic surgery, the focus should not solely be on the condylar changes, but importantly, on the change of the entire TMJ, involving all three adaptive processes. No studies have been identified, which propose a validated holistic approach for 3D assessment of the TMJ including all three adaptive processes. However, the accuracy of the 3D assessment of condylar changes from CBCT has been studied previously and the assessment was shown to be highly accurate for both volumetric and surface discrepancy measurements [23]. Similarly, superimposition on the cranial base and the mandibular ramus has shown high accuracy [19,20]. Hence, the purpose of the present study was to propose and assess the reliability of a semi-automatic approach for holistic 3D assessment of the TMJ from CBCT following orthognathic surgery, including condylar and glenoid fossa changes, as well as the interrelated positional changes. The null hypothesis was: H0: it is not possible to develop a reliable holistic approach for 3D assessment of the TMJ. The approach was deemed reliable if the intra-class correlation coefficient (ICC) was good (>0.60), according to Cicchetti [24].

2. Materials and Methods

2.1. Study Sample

This study was based on pre- and postsurgical (two years) CBCT-scans from a study sample diagnosed with maxillary and/or mandibular growth disturbances, who underwent a combined bilateral sagittal split osteotomy (BSSO) and Le Fort I procedure at the Department of Oral and Maxillofacial Surgery, University Hospital of Southern Denmark, Esbjerg, Denmark. The study sample was selected, such that half of the subjects were diagnosed with postoperative condylar resorption and the other half without.
Inclusion criteria: age range of 18–65 years; diagnosis indicating a combined BSSO and Le Fort I osteotomy; availability of patients’ pre- and postoperative (two years) CBCT scans. Exclusion criteria: previous history of oral and maxillofacial surgery, presence of craniofacial anomaly syndrome, rheumatoid arthritis, or previous trauma.

2.2. Image Acquisition

The CBCT-images were acquired using an i-CAT scanner, version 17–19 (Imaging Sciences International, Hatfield, PA: 120 kVp; 5 mA; 7 s exposure time; Field-of-View (FOV) = 23 × 17.8 cm (768 × 768 × 576 voxels)); isotropic voxels of 0.30 mm. The subjects were scanned in an upright natural head position. The condyles were seated in centric relation and the preoperative occlusion was fixated by a wax-bite, however, without a wax-bite at the postoperative CBCT scanning acquisition. The CBCT data were exported in DICOM format and imported into Mimics® 24 (Materialise NV, Leuven, Belgium).

2.3. Surgical Technique

The surgery was performed with general anesthesia as a mandible-first procedure. After BSSO, the distal mandibular segment was positioned using an intermediate splint and was fixated bilaterally to each proximal segment with two 2.0 four hole osteosynthesis plates (KLS Martin, Tuttlingen, Germany). Following the Le Fort I osteotomy, the maxilla was segmented into three pieces, in segmental cases, and the tooth-bearing segments were positioned into the final occlusion without the use of a splint, and subsequently fixated using two 2.0 Y-plates anteriorly and two L-plates posteriorly (KLS Martin, Tuttlingen, Germany). All osteotomy sites were grafted.

2.4. Holistic Assessment of the TMJ

The holistic assessment of the TMJ including all three adaptive processes (Figure 1) was performed using Mimics® 24.0 and 3-matic® 16.0 (Materialise NV, Leuven, Belgium). The workflow was automated using Python scripting, Python 3.8 (Python Software Foundation, Fredericksburg, Virginia), and semi-automation was achieved using superimposition techniques as shown in Figure 2 and described in the following.

2.4.1. Assessment of Condylar Changes

The pre- and postoperative mandibles were semi-automatically segmented according to Holte et al. [20]. Subsequently, the segmentation of the mandibular condyle was refined and registration was automatically performed to align the right and left pre- and postoperative rami, separately, as proposed in the protocol validated by Verhelst et al. [23]. Surface-based registration was used for the alignment, which was shown to be more accurate and reliable than VBR for the assessment of mandibular condyle remodeling [20].
The C-plane, centered in the pre-operative mandibular notch (C-point) and parallel to the Frankfurt horizontal plane, as defined by Xi et al. [25], was used to isolate the pre- and postoperative condyles from the rami. For spatial analysis, the condylar head was spatially divided into four sub-regions: anterior/posterior-lateral and medial condylar head using two cutting planes: (1) a plane going through the lateral and medial condylar poles perpendicular to the horizontal Frankfurt plane, and (2) an orthogonal mid-plane defined by the lateral and medial condylar poles. Postoperative change of the condyle was represented by the volumetric change and by color-coded distance maps [14], quantified by the mean surface distance of the condylar head and neck, and the defined four sub-regions of the condylar head (Figure 1).

2.4.2. Assessment of Glenoid Fossa Changes

The pre- and postoperative crania were semi-automatically segmented according to Holte et al. [4] For alignment of the segmented pre- and postoperative crania, the postoperative CBCT scan was registered to the preoperative CBCT scan by VBR using the anterior cranial base, zygomatic arches, and forehead as the volume of interest unaffected by the surgery [19]. Next, a curve was manually traced engulfing the preoperative glenoid fossa, which was automatically attracted and attached to the surface of the postoperative skull, defining the postoperative glenoid fossa (Figure 1). A three-dimensional assessment of glenoid fossa changes was performed according to Holte et al. [4]. However, instead of applying the midsagittal and coronal plane for spatial division of the glenoid fossa through its center of gravity [4], the glenoid fossa was divided analogously to the division of the condylar head into four sub-regions for spatial analysis using the two previously defined cutting-planes. Glenoid fossa surface discrepancies were represented by color-coded distance maps, [14] and quantified by the root mean square (RMS) surface distance of the total glenoid fossa and the four defined fossa sub-regions (Figure 1) [4].

2.4.3. Assessment of Interrelated Positional Change in the Joint Space

The 3D assessment of interrelated positional change in the joint space was calculated as the pre- to postoperative change in the minimum distance between the mandibular condyle and the glenoid fossa for each of the four sub-regions (Figure 1).

2.5. Statistical Analysis

Statistical analysis of the data was performed in STATA® 16.1 (StataCorp, College Station, TX, USA). A sample size calculation for a one-sample correlation study (ICC > 0.6, power = 0.8, alpha = 0.05, and two raters) was performed. To evaluate the reliability, two observers independently performed the assessment (MBH, HS). The inter-observer reliability of the measurements was summarized using the mean absolute difference (MAD) and standard deviation (SD), similar to related reliability studies [9,10,11,12,20,23]. Intraclass correlation coefficients at a 95% confidence interval on measurements of the two observers were calculated for single measurements using a one-way random effect model. Bland-Altman plots were produced to evaluate observer agreement with 95% limits of agreement.

3. Results

The statistical sample size calculation resulted in a required sample size of n = 20 fossae to obtain a statistical power of 0.8. Hence, ten post-pubertal patients in a database (March 2012–November 2017) who met the inclusion- and exclusion criteria, were included. Nine female; one male; mean age 25.6 ± 6.9 years; skeletal class II malocclusion with mandibular retrognathia (eight subjects); maxillary retrognathia (two subjects); anterior open bite (nine subjects); mandibular asymmetry (three subjects); treated with maxillomandibular advancement (ten subjects); maxillary expansion (ten subjects) and with genioplasty (two subjects).
The inter-observer ICCs MADs and SDs are presented in Table 1 and Table 2. The inter-observer ICCs for the volumetric and mean surface condylar measurements, glenoid fossa RMS distance measurements, and change in minimum joint space distance measurements were all excellent: 1.00, 0.99, 0.82 and 0.91, respectively. The inter-observer MAD (SD) were 1.68% (1.58), 0.10 mm (0.14), 0.06 mm (0.04) and 0.17 mm (0.16), respectively. For 3D assessment of the condylar head and neck the ICCs were excellent, (0.99–1.00) and (0.96–0.98), with inter-observer volumetric MAD (SD) values of 2.71% (2.73) and 2.26% (2.04), respectively, and with inter-observer mean surface-distance MAD (SD) values of 0.17 mm (0.28) and 0.09 mm (0.12), respectively.
Spatial division of the condylar head, the glenoid fossa and the joint space into the four sub-regions, resulted in an increased range of the inter-observer ICCs from good to excellent, (0.96–1.00), (0.71–0.86) and (0.89–0.97), and an increased range of the MAD (SD) values, (2.32% (2.13)–5.01% (3.85)), (0.05 mm (0.05)–0.08 mm (0.06)) and (0.12 mm (0.09)–0.19 mm (0.18)), respectively.
Figure 3 shows Bland–Altman plots of the inter-observer agreement on the measurements. For Bland-Altman plots of the inter-observer agreement on the measurements of the spatial divided condylar head, glenoid fossa and joint space, please refer to Supplementary Materials. The plots show a high degree of consistency in form of low biases, low limits of agreement and few outliers.

4. Discussion

The purpose of the present study was to propose and assess the reliability of a semi-automatic approach for holistic 3D assessment of the TMJ from CBCT following orthognathic surgery, including condylar and glenoid fossa changes, as well as the interrelated positional changes. The null hypothesis was: H0: it is not possible to develop a reliable holistic approach for 3D assessment of the TMJ. The approach was deemed reliable if the ICC was good (>0.60), according to Cicchetti [24].
The null hypothesis was rejected by the assessment, showing that the proposed approach has good to excellent inter-observer reliability. When assessing the TMJ without spatial division, the inter-observer reliability was excellent for 3D assessment of all three adaptive processes. When spatially dividing the TMJ into the four sub-regions, the inter-observer reliability remained excellent for 3D assessment of condylar and the interrelated positional changes, whilst the 3D assessment of glenoid fossa changes was excellent for the anterior regions and good for the posterior regions. Both mean- and RMS surface distances were reported in order to demonstrate the reliability of both measures. The minimum or shortest distance has generally been applied in radiographic studies of the TMJ space for decades [26].
The Bland-Altman plots (Figure 3) showed a high degree of observer agreement with only a single outlier in the measurement of condylar and joint space changes. This outlier was caused by a modest inter-observer discrepancy in the assessment of a subject with a high degree of condylar resorption. This tendency seems to be more evident for the mean surface distance measurements of condylar changes, while assessment of subject little to no condylar resorption showed a very high inter-observer agreement. Similarly, the Bland-Altman plots of the spatial 3D assessment (Supplementary Materials) showed a high degree of inter-observer agreement, however, a few outliers were present in the inter-observer measurements. These outliers were mainly caused by an inter-observer discrepancy in the segmentation and the registration results.
The roof of the glenoid fossa is very thin and can cause discontinuities in 3D CBCT scans due to low contrast, spatial resolution, noise, and artefacts, making it difficult to accurately segment and evaluate the fossa [27]. The present approach applied a semi-automatic method for segmentation and 3D reconstruction of the glenoid fossa [4]. Often the glenoid fossa cannot be completely segmented using global image segmentation techniques, and manual segmentation is generally time-consuming and subject to operator variation. For the assessment of relatively small morphovolumetrical changes, such variation affects the reliability of the assessment.
The present study proposed a reliable holistic approach, which can be applied in future clinical studies for semi-automatic 3D evaluation of the TMJ, involving spatial assessment of morphovolumetrical changes, e.g., following orthodontic treatment, orthognathic surgery, trauma, general disease such as rheumatoid arthritis, or to study mandibular growth or asymmetry. The assessment focused on the reliability of the approach. The reliability of the present 3D assessment of condylar changes is in line with the related studies by Verhelst et al. [23] and Xi et al. [28], who reported similar ICC and MAD values.
The accuracy of the 3D assessment of condylar changes from CBCT has been studied previously and the assessment was shown to be highly accurate for both volumetric and surface discrepancy measurements [23]. Similarly, superimposition on the cranial base and the mandibular ramus has shown high accuracy [19,20]. In the study by Holte et al. [20], it was shown that SBR on the mandibular ramus was accurate and reliable for long-term 3D assessment of condylar remodeling following orthognathic surgery, and the measurement error was considered clinically irrelevant. On the contrary, manual landmark-based 3D cephalometric analysis was shown to accumulate landmark identification errors, ranging from 0.02 to 2.47 mm [29,30,31]. Such errors influence the quality of the measurements and make the assessment costly in terms of manual processing time.
Holte et al. also compared the performance of SBR and VBR on the mandibular ramus for long-term 3D assessment of condylar remodeling following orthognathic surgery [20]. The comparative study showed that SBR was more accurate and reliable than VBR using the reference structures proposed by Verhelst et al. [23]. However, it was concluded that the performance of the registration on the mandibular ramus may have been compromised by inappropriate reference structures proposed in the literature, which seem to alter due to remodeling, and may have influenced the registration [20]. Related studies comparing the performance of VBR and SBR for assessment of the outcome of orthognathic surgery [12] and in adult orthodontic subjects [18] found the two registration methods to be accurate and reliable. Voxel-based registration was associated with less variability [12] and was more efficient than SBR [18]. However, the statistical differences between the two registration methods were insignificant and were unlikely to have any clinical significance [12,18].
Several methods have been proposed for the 3D assessment of condylar changes following orthognathic surgery [1,2,3,23,28,32]. However, most of the previous studies on resorption and remodeling of the mandibular fossa relied on a two-dimensional (2D) assessment of lateral cephalograms [15]. Before the introduction of CBCT, 2D transcranial radiography was the standardized imaging modality for studying TMJ [26,33]. However, it is inadequate to describe the complex 3D process of the TMJ remodelling caused by treatment [5,6]. The risk of multiple errors with 2D imaging exists, such as distortion-, magnification-, positioning errors (patient) and overlapping anatomic structures, making landmark identification difficult and cephalometry insufficient to explain the complex 3D process. However, it should be noted that the 3D analysis of the TMJ from CBCT is limited to the assessment of osseous changes, while magnetic resonance imaging enables the assessment of the soft tissue components of the TMJ, i.e., the articular disc, the synovial membrane and the lateral pterygoid muscle [33,34,35].
Recently, fully automated image segmentation of the craniomaxillofacial bones from CBCT has been proposed [36,37]. These studies applied artificial intelligence, more specifically deep learning by neural networks, thereby eliminating the need of manual image segmentation and operator variability [36,37]. Artificial intelligence has been applied for automatic segmentation of the mandibular ramus and condyle [38]. The authors concluded that their findings suggest that CBCT image segmentation of the mandibular ramus and condyle from different clinical centers can be analyzed effectively [38]. Hence, in future studies the semi-automatic segmentation of the mandibular condyle applied in the present study may be substituted with a fully automatic segmentation using artificial intelligence.
To the best of the authors’ knowledge, this is the first study to propose and assess the reliability of a holistic 3D assessment of the TMJ including all three adaptive processes that are believed to contribute to the position of the mandible [21]. Optionally, the proposed method can be combined with the approach by Holte et al [9] in order to include the 3D translational and rotational condylar displacement with six degrees of freedom. The application of a holistic approach is important in order to achieve a more complete understanding, e.g., of the postoperative stability and functionality following orthognathic surgery. Validation of the approach is important in order to obtain accurate and reliable measurements. A semi-automatic implementation of the 3D assessment enables standardization of the measurements, which is important in order to couple and compare results for more conclusive evidence and recommendations based on future meta-analyses. The importance of automation and standardization was concluded in the recent systematic literature review and meta analysis on 3D accuracy and stability of personalized implants in orthognathic surgery, where great heterogeneity was observed in the assessment methods applied in the reviewed literature [39]. Hence, a meta-analysis could only be performed on a sparse subset of the identified studies [39].

5. Conclusions

The proposed semi-automatic approach was shown to have good to excellent reliability. It facilitates a holistic 3D quantification and assessment of the TMJ from CBCT, including all three adaptive processes in the TMJ, which are believed to contribute to the position of the mandible: (1) adaptive condylar changes, (2) glenoid fossa changes, and (3) condylar positional changes within the fossa. Application of a standardized holistic approach is important in order to study the TMJ, e.g., to achieve a more complete understanding of its response to orthodontic treatment, orthognathic surgery, trauma, general disease such as rheumatoid arthritis, or to study mandibular growth or asymmetry.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/jpm13020343/s1, Additional Bland-Altman plots.

Author Contributions

Conceptualization, M.B.H.; methodology, M.B.H. and E.M.P.; software, M.B.H.; validation, M.B.H. and H.S.; formal analysis, M.B.H.; investigation, M.B.H.; data curation, M.B.H.; writing—original draft preparation, M.B.H.; writing—review and editing, M.B.H., H.S. and E.M.P.; visualization, M.B.H.; supervision, M.B.H. and E.M.P.; project administration, M.B.H. and E.M.P. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Permission was granted by the Institutional Ethics Committee—University Hospital of Southern Denmark, Esbjerg (21/33871). As the study comprised retrospective material, none of the subjects were exposed to any extra radiation and no extra examination has been performed to acquire additional information. The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of University Hospital of Southern Denmark, Esbjerg (21/33871, 14 June 2021).

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Acknowledgments

The authors thank senior surgeons Janne Ingerslev and Jens Jørgen Thorn, Department of Oral and Maxillofacial Surgery, University Hospital of Southern Denmark, Esbjerg, for clinical counsel; and Gabriele Berg-Beckhoff, Department of Public Health, Faculty of Health Sciences, University of Southern Denmark, Esbjerg, for statistical counsel.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Verhelst, P.J.; Verstraete, L.; Shaheen, E.; Shujaat, S.; Darche, V.; Jacobs, R.; Swennen, G.; Politis, C. Three-dimensional cone beam computed tomography analysis protocols for condylar remodelling following orthognathic surgery: A systematic review. Int. J. Oral Maxillofac. Surg. 2020, 49, 207–217. [Google Scholar] [CrossRef] [Green Version]
  2. Vandeput, A.S.; Verhelst, P.J.; Jacobs, R.; Shaheen, E.; Swennen, G.; Politis, C. Condylar changes after orthognathic surgery for class III dentofacial deformity: A systematic review. Int. J. Oral Maxillofac. Surg. 2019, 48, 193–202. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  3. Mousoulea, S.; Kloukos, D.; Sampaziotis, D.; Vogiatzi, T.; Eliades, T. Condylar resorption in orthognathic patients after mandibular bilateral sagittal split osteotomy: A systematic review. Eur. J. Orthod. 2017, 39, 294–309. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  4. Holte, M.B.; Nielsen, T.W.; Saederup, H.; Pinholt, E.M. Does the Glenoid Fossa Change Following Orthognathic Surgery? J. Oral Maxillofac. Surg. 2022, 80, 1007–1017. [Google Scholar] [CrossRef] [PubMed]
  5. LeCornu, M.; Cevidanes, L.H.; Zhu, H.; Wu, C.D.; Larson, B.; Nguyen, T. Three-dimensional treatment outcomes in Class II patients treated with the Herbst appliance: A pilot study. Am. J. Orthod. Dentofacial. Orthop. 2013, 144, 818–830. [Google Scholar] [CrossRef] [Green Version]
  6. Atresh, A.; Cevidanes, L.H.S.; Yatabe, M.; Muniz, L.; Nguyen, T.; Larson, B.; Manton, D.J.; Schneider, P.M. Three-dimensional treatment outcomes in Class II patients with different vertical facial patterns treated with the Herbst appliance. Am. J. Orthod. Dentofacial. Orthop. 2018, 154, 238–248 e231. [Google Scholar] [CrossRef]
  7. Jung, J.; Kim, J.H.; Lee, J.W.; Ohe, J.Y.; Choi, B.J. Three-dimensional volumetric analysis of condylar head and glenoid cavity after mandibular advancement. J. Craniomaxillofac. Surg. 2018, 46, 1470–1475. [Google Scholar] [CrossRef] [PubMed]
  8. Kim, J.Y.; Kim, B.J.; Park, K.H.; Huh, J.K. Comparison of volume and position of the temporomandibular joint structures in patients with mandibular asymmetry. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. 2016, 122, 772–780. [Google Scholar] [CrossRef]
  9. Holte, M.B.; Diaconu, A.; Ingerslev, J.; Thorn, J.J.; Pinholt, E.M. Virtual surgical analysis: Long-term cone beam computed tomography stability assessment of segmental bimaxillary surgery. Int. J. Oral Maxillofac. Surg. 2022, 51, 1188–1196. [Google Scholar] [CrossRef]
  10. Holte, M.B.; Diaconu, A.; Ingerslev, J.; Thorn, J.J.; Pinholt, E.M. Virtual Analysis of Segmental Bimaxillary Surgery: A Validation Study. J. Oral Maxillofac. Surg. 2021, 79, 2320–2333. [Google Scholar] [CrossRef]
  11. Diaconu, A.; Holte, M.B.; Cattaneo, P.M.; Pinholt, E.M. A semi-automatic approach for longitudinal 3D upper airway analysis using voxel-based registration. Dentomaxillofac. Radiol. 2022, 51, 20210253. [Google Scholar] [CrossRef] [PubMed]
  12. Almukhtar, A.; Ju, X.; Khambay, B.; McDonald, J.; Ayoub, A. Comparison of the accuracy of voxel based registration and surface based registration for 3D assessment of surgical change following orthognathic surgery. PLoS ONE 2014, 9, e93402. [Google Scholar] [CrossRef] [Green Version]
  13. Maes, F.; Collignon, A.; Vandermeulen, D.; Marchal, G.; Suetens, P. Multimodality image registration by maximization of mutual information. IEEE Trans. Med. Imaging 1997, 16, 187–198. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  14. Cevidanes, L.H.; Bailey, L.J.; Tucker, G.R., Jr.; Styner, M.A.; Mol, A.; Phillips, C.L.; Proffit, W.R.; Turvey, T. Superimposition of 3D cone-beam CT models of orthognathic surgery patients. Dentomaxillofac. Radiol. 2005, 34, 369–375. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  15. Kula, K.; Ghoneima, A. Cephalometry in Orthodontics: 2D and 3D; Quintessence Publishing Company: Batavia, IL, USA, 2020. [Google Scholar]
  16. Xi, T.; van Luijn, R.; Baan, F.; Schreurs, R.; de Koning, M.; Berge, S.; Maal, T. Landmark-Based Versus Voxel-Based 3-Dimensional Quantitative Analysis of Bimaxillary Osteotomies: A Comparative Study. J. Oral Maxillofac. Surg. 2020, 78, e468.e1–e468.e10. [Google Scholar] [CrossRef]
  17. Andriola, F.O.; Haas Junior, O.L.; Guijarro-Martinez, R.; Hernandez-Alfaro, F.; Oliveira, R.B.; Pagnoncelli, R.M.; Swennen, G.R. Computed tomography imaging superimposition protocols to assess outcomes in orthognathic surgery: A systematic review with comprehensive recommendations. Dentomaxillofac. Radiol. 2022, 51, 20210340. [Google Scholar] [CrossRef]
  18. Han, G.; Li, J.; Wang, S.; Wang, L.; Zhou, Y.; Liu, Y. A comparison of voxel- and surface-based cone-beam computed tomography mandibular superimposition in adult orthodontic patients. J. Int. Med. Res. 2021, 49, 300060520982708. [Google Scholar] [CrossRef]
  19. Nada, R.M.; Maal, T.J.; Breuning, K.H.; Berge, S.J.; Mostafa, Y.A.; Kuijpers-Jagtman, A.M. Accuracy and reproducibility of voxel based superimposition of cone beam computed tomography models on the anterior cranial base and the zygomatic arches. PLoS ONE 2011, 6, e16520. [Google Scholar] [CrossRef] [Green Version]
  20. Holte, M.B.; Saederup, H.; Pinholt, E.M. Comparison of surface- and voxel-based registration on the mandibular ramus for long-term three-dimensional assessment of condylar remodelling following orthognathic surgery. Dentomaxillofac. Radiol. 2022, 51, 20210499. [Google Scholar] [CrossRef]
  21. Ruf, S.; Pancherz, H. Temporomandibular joint growth adaptation in Herbst treatment: A prospective magnetic resonance imaging and cephalometric roentgenographic study. Eur. J. Orthod. 1998, 20, 375–388. [Google Scholar] [CrossRef] [Green Version]
  22. Woodside, D.G.; Metaxas, A.; Altuna, G. The influence of functional appliance therapy on glenoid fossa remodeling. Am. J. Orthod. Dentofacial. Orthop. 1987, 92, 181–198. [Google Scholar] [CrossRef]
  23. Verhelst, P.J.; Shaheen, E.; de Faria Vasconcelos, K.; Van der Cruyssen, F.; Shujaat, S.; Coudyzer, W.; Salmon, B.; Swennen, G.; Politis, C.; Jacobs, R. Validation of a 3D CBCT-based protocol for the follow-up of mandibular condyle remodeling. Dentomaxillofac. Radiol. 2020, 49, 20190364. [Google Scholar] [CrossRef] [PubMed]
  24. Cicchetti, D.V. Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychol. Assess. 1994, 6, 284. [Google Scholar] [CrossRef]
  25. Xi, T.; van Luijn, R.; Baan, F.; Schreurs, R.; de Koning, M.; Berge, S.; Maal, T. Three-dimensional analysis of condylar remodeling and skeletal relapse following bimaxillary surgery: A 2-year follow-up study. J. Craniomaxillofac. Surg. 2017, 45, 1311–1318. [Google Scholar] [CrossRef]
  26. Lindahl, L.; Hollender, L. Condylar fractures of the mandible. II. a radiographic study of remodeling processes in the temporomandibular joint. Int. J. Oral Surg. 1977, 6, 153–165. [Google Scholar] [CrossRef] [PubMed]
  27. Khojastepour, L.; Haghnegahdar, A.; Eisazadeh, M.; Bahreini, M. Comparison between Glenoid Fossa Roof Thickness in TMD and non-TMD Patients, a CBCT Study. J. Dent. 2019, 20, 165–170. [Google Scholar] [CrossRef]
  28. Xi, T.; Schreurs, R.; Heerink, W.J.; Berge, S.J.; Maal, T.J. A novel region-growing based semi-automatic segmentation protocol for three-dimensional condylar reconstruction using cone beam computed tomography (CBCT). PLoS ONE 2014, 9, e111126. [Google Scholar] [CrossRef]
  29. Katkar, R.A.; Kummet, C.; Dawson, D.; Moreno Uribe, L.; Allareddy, V.; Finkelstein, M.; Ruprecht, A. Comparison of observer reliability of three-dimensional cephalometric landmark identification on subject images from Galileos and i-CAT cone beam CT. Dentomaxillofac. Radiol. 2013, 42, 20130059. [Google Scholar] [CrossRef] [Green Version]
  30. Titiz, I.; Laubinger, M.; Keller, T.; Hertrich, K.; Hirschfelder, U. Repeatability and reproducibility of landmarks--a three-dimensional computed tomography study. Eur. J. Orthod. 2012, 34, 276–286. [Google Scholar] [CrossRef] [Green Version]
  31. Swennen, G.R.; Schutyser, F.; Barth, E.L.; De Groeve, P.; De Mey, A. A new method of 3-D cephalometry Part I: The anatomic Cartesian 3-D reference system. J. Craniofac. Surg. 2006, 17, 314–325. [Google Scholar] [CrossRef]
  32. Schilling, J.; Gomes, L.C.; Benavides, E.; Nguyen, T.; Paniagua, B.; Styner, M.; Boen, V.; Goncalves, J.R.; Cevidanes, L.H. Regional 3D superimposition to assess temporomandibular joint condylar morphology. Dentomaxillofac. Radiol. 2014, 43, 20130273. [Google Scholar] [CrossRef] [PubMed]
  33. Talmaceanu, D.; Lenghel, L.M.; Bolog, N.; Hedesiu, M.; Buduru, S.; Rotar, H.; Baciut, M.; Baciut, G. Imaging modalities for temporomandibular joint disorders: An update. Clujul. Med. 2018, 91, 280–287. [Google Scholar] [CrossRef] [PubMed]
  34. Bedran, L.M.; Dos Santos, A. Changes in temporomandibular joint anatomy, changes in condylar translation, and their relationship with disc displacement: Magnetic resonance imaging study. Radiol. Bras. 2019, 52, 85–91. [Google Scholar] [CrossRef]
  35. Ruf, S.; Pancherz, H. Temporomandibular joint remodeling in adolescents and young adults during Herbst treatment: A prospective longitudinal magnetic resonance imaging and cephalometric radiographic investigation. Am. J. Orthod. Dentofacial. Orthop. 1999, 115, 607–618. [Google Scholar] [CrossRef] [PubMed]
  36. Gillot, M.; Baquero, B.; Le, C.; Deleat-Besson, R.; Bianchi, J.; Ruellas, A.; Gurgel, M.; Yatabe, M.; Al Turkestani, N.; Najarian, K.; et al. Automatic multi-anatomical skull structure segmentation of cone-beam computed tomography scans using 3D UNETR. PLoS ONE 2022, 17, e0275033. [Google Scholar] [CrossRef] [PubMed]
  37. Zhang, J.; Liu, M.; Wang, L.; Chen, S.; Yuan, P.; Li, J.; Shen, S.G.; Tang, Z.; Chen, K.C.; Xia, J.J.; et al. Context-guided fully convolutional networks for joint craniomaxillofacial bone segmentation and landmark digitization. Med. Image Anal. 2020, 60, 101621. [Google Scholar] [CrossRef] [PubMed]
  38. Le, C.; Deleat-Besson, R.; Prieto, J.; Brosset, S.; Dumont, M.; Zhang, W.; Cevidanes, L.; Bianchi, J.; Ruellas, A.; Gomes, L.; et al. Automatic Segmentation of Mandibular Ramus and Condyles. Annu. Int. Conf IEEE Eng. Med. Biol. Soc. 2021, 2021, 2952–2955. [Google Scholar] [CrossRef]
  39. Diaconu, A.; Holte, M.B.; Berg-Beckhoff, G.; Pinholt, E.M. Three-Dimensional Accuracy and Stability of Personalized Implants in Orthognathic Surgery: A Systematic Review and a Meta-Analysis. J. Pers. Med. 2023, 13, 125. [Google Scholar] [CrossRef] [PubMed]
Jpm 13 00343 g001 550
Figure 1. Three-dimensional assessment of morphovolumetrical changes of the TMJ, including all three adaptive processes, which are believed to contribute to the position of the mandible: (a) Volumetric condylar changes; (b) Three-dimensional fossa changes visualized by color-coded distance maps; (c) Minimum 3D joint space distances.
Figure 1. Three-dimensional assessment of morphovolumetrical changes of the TMJ, including all three adaptive processes, which are believed to contribute to the position of the mandible: (a) Volumetric condylar changes; (b) Three-dimensional fossa changes visualized by color-coded distance maps; (c) Minimum 3D joint space distances.
Jpm 13 00343 g001
Jpm 13 00343 g002 550
Figure 2. Semi-automation of the 3D assessment using superimposition: (a) Voxel-based registration on the anterior cranial base; (b) Resulting alignment of the pre- and postoperative cranial 3D models; (c) Surface-based registration on the mandibular rami using the stable reference structure proposed by Verhelst et al. [23].
Figure 2. Semi-automation of the 3D assessment using superimposition: (a) Voxel-based registration on the anterior cranial base; (b) Resulting alignment of the pre- and postoperative cranial 3D models; (c) Surface-based registration on the mandibular rami using the stable reference structure proposed by Verhelst et al. [23].
Jpm 13 00343 g002
Jpm 13 00343 g003 550
Figure 3. Bland-Altman plots of inter-observer agreement on the 3D TMJ assessments of: (a) Volumetric condylar changes; (b) condylar changes measured by mean surface distance; (c) fossa changes measured by root-mean-squre (RMS) surface distance; (d) the minimum 3D joint space distance.
Figure 3. Bland-Altman plots of inter-observer agreement on the 3D TMJ assessments of: (a) Volumetric condylar changes; (b) condylar changes measured by mean surface distance; (c) fossa changes measured by root-mean-squre (RMS) surface distance; (d) the minimum 3D joint space distance.
Jpm 13 00343 g003
Table
Table 1. Intraclass correlation coefficients (ICC), mean absolute differences (MAD) and standard deviations (SD for inter-observer measurements of volumetric condylar remodeling.
Table 1. Intraclass correlation coefficients (ICC), mean absolute differences (MAD) and standard deviations (SD for inter-observer measurements of volumetric condylar remodeling.
Volumetric Change (%)
ICCMAD (SD)
Condyle1.001.68 (1.58)
  Condylar neck0.982.26 (2.04)
  Condylar head1.002.71 (2.73)
     Anterior-lateral0.993.88 (4.69)
     Anterior-medial0.995.01 (3.85)
     Posterior-lateral1.002.32 (2.13)
     Posterior-medial0.993.29 (2.99)
Table
Table 2. Intraclass correlation coefficients (ICC), mean absolute differences (MAD) and standard deviations (SD for inter-observer distance measurements of condylar and glenoid fossa remodeling and changes to the temporomandibular joint space.
Table 2. Intraclass correlation coefficients (ICC), mean absolute differences (MAD) and standard deviations (SD for inter-observer distance measurements of condylar and glenoid fossa remodeling and changes to the temporomandibular joint space.
Mean Surface Distance (mm)
ICCMAD (SD)
Condyle0.990.10 (0.14)
  Condylar neck0.960.09 (0.12)
  Condylar head0.990.17 (0.28)
     Anterior-lateral0.990.19 (0.28)
     Anterior-medial0.990.15 (0.20)
     Posterior-lateral0.960.25 (0.46)
     Posterior-medial0.990.14 (0.19)
RMS surface distance (mm)
ICCMAD (SD)
Glenoid fossa0.820.06 (0.04)
  Anterior-lateral0.860.05 (0.05)
  Anterior-medial0.860.06 (0.05)
  Posterior-lateral0.710.08 (0.06)
  Posterior-medial0.770.07 (0.08)
Change in minimum joint space distance (mm)
ICCMAD (SD)
TMJ joint space0.910.17 (0.16)
  Anterior-lateral0.970.12 (0.09)
  Anterior-medial0.890.19 (0.16)
  Posterior-lateral0.970.15 (0.10)
  Posterior-medial0.920.19 (0.18)
RMS: Root mean square.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Holte, M.B.; Sæderup, H.; Pinholt, E.M. A Semi-Automatic Approach for Holistic 3D Assessment of Temporomandibular Joint Changes. J. Pers. Med. 2023, 13, 343. https://doi.org/10.3390/jpm13020343

AMA Style

Holte MB, Sæderup H, Pinholt EM. A Semi-Automatic Approach for Holistic 3D Assessment of Temporomandibular Joint Changes. Journal of Personalized Medicine. 2023; 13(2):343. https://doi.org/10.3390/jpm13020343

Chicago/Turabian Style

Holte, Michael Boelstoft, Henrik Sæderup, and Else Marie Pinholt. 2023. "A Semi-Automatic Approach for Holistic 3D Assessment of Temporomandibular Joint Changes" Journal of Personalized Medicine 13, no. 2: 343. https://doi.org/10.3390/jpm13020343

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop