Introduction
Materials and methods
Protocol & registration
Eligibility criteria
Criteria | Specification |
---|---|
Focus question | What is the best technique to construct a mid-sagittal reference plane for the estimation of facial asymmetry? |
Population | Patients with clinically diagnosed facial asymmetry or craniofacial deformity |
Intervention | The use of any three-dimensional (3D) technology-based tool, device, software, or intervention for the estimation of the mid-sagittal plane |
Comparator/control | Different midsagittal plane construction methods; different types of midsagittal planes; asymmetry or normal controls; different asymmetry quantification methods |
Outcomes | a) Effectiveness of the technique b) Ease of clinical applicability |
Search strategy | Search (3-dimensional OR 3-D OR 3D OR three-dimensional OR mesh (three-dimensional)) AND (midsagittal OR midsagittal OR (midsagittal plane) OR MSP OR MRP OR (midsagittal reference plane)) AND ((facial asymmetry) OR (asymmetric face) OR (asymmetrical face) OR mesh (facial asymmetry)) |
Information sources and literature search
Study selection
Data extraction and outcomes of interest
Quality analysis
Level of evidence
Results
Study selection
Study characteristics
Author (year) | Study type | Sample size | Sex (M/F) | Mean age/age range (years) | Ethnicity | Skeletal discrepancy | Asymmetry criteria |
---|---|---|---|---|---|---|---|
Wong (2005) | Computational | 1 | F | 17 | Asian | Class III | n/r |
Hartmann (2007) | Computational | n/r | n/r | n/r | Caucasian | n/r | n/r |
AlHadidi (2011) | Observational | 50 | n/r | n/r | Caucasian | n/r | Chin > 2 mm Presence of cant of occlusal plane |
Kim (2011) | Retrospective | 102 | F | 21.551 ± 2.644/ 20–29 | Asian | Class III | n/r |
Baek (2012) | Retrospective | 43 9, Class I 4, Class II 30, Class III | 18/25 | 24.3 ± 4.4 | Asian | Class I, Class II, Class III | n/r |
Damstra (2012) | Observational | 14 skulls 5, Asymmetry 9, Symmetry | n/r | n/r | Caucasian | n/r | Me > 4 mm |
Berssenbrügge (2014) | Computational | 50 faces | 22 / 28 | 20–32 | Caucasian | n/r | n/r |
Wong (2014) | Retrospective | 20 | 13 / 7 | 24.3 | Asian | Condylar hyperplasia/hypoplasia; Hemifacial microsomia; Condylar tumour; Trauma related | Me > 4 mm/ 4° |
Gateno (2015) | Experimental | 1 | M | n/r | Caucasian | n/r | n/r |
Kim (2015) | Retrospective | 24 | 12 / 12 | 22.5/ 18.2–29.7 | Asian | n/r | Normal, 0 mm ≤ Me < 2 mm Mild, 2 mm ≤ Me < 4 mm Moderate, 4 mm ≤ Me < 8 mm Severe, 8 mm ≤ Me |
Ryu (2015) | Observational | 85 | Asian | Class III | Me < 2 mm Me > 3 mm | ||
30, Control | 15/15 | 24.30 ± 4.14 | |||||
55, Asymmetry: | |||||||
Hyperdivergent Hypodivergent | 13/15 15/12 | 26.34 ± 3.14 28.96 ± 5.27 | |||||
Lee (2016) | Retrospective | 35 | Asian | n/r | Me < 2 mm Me ≥ 4 mm | ||
15, Symmetry | 7/8 | 22.3 ± 3.3 | |||||
20, Asymmetry | 12/8 | 21.9 ± 3.4 | |||||
Shin (2016) | Retrospective | 69 | 36 / 33 | 23.0 ± 4.1 | Asian | Me > 3.6 ± 2.3 mm | |
10, Class I | 4 / 6 | M, 22.9 ± 4.1 F, 23.0 ± 4.2 | Class I, | ||||
12, Class II | 5 / 7 | Class II | |||||
47, Class III | 24 / 23 | Class III | |||||
Song (2016) | Retrospective | 29 | Asian | Class III | Me > 4 mm | ||
16, CS | 10/6 | 21.8 ± 2.2 | |||||
13, POGS | 5/8 | 21.2 ± 4.3 | |||||
Sangln An (2017) | Observational | 30 | n/r | 25.7 ± 6.03/19–43 | Asian | n/r | n/r |
Dobai (2018) | Retrospective | 60 | Caucasian | n/r | n/r | ||
30, Group I | 11 / 19 | 18–30 | |||||
30, Group II | 12/ 18 | 20–28 | |||||
Economou (2018) | Observational | 21 | 7/14 | 13.5 | Caucasian | Juvenile Idiopathic Arthritis | n/r |
Jajoo (2018) | Computational | 20 CT skull models | n/r | n/r | Caucasian | Horizontal/ Vertical condylar hyperplasia; Type 1/ Type 2 Hemifacial microsomia | n/r |
Oh (2018) | Retrospective | 60 | Asian | Me > 2º | |||
30, Asymmetry | 15/15 | 23.2 ± 3.8 | Class I, Class III | ||||
30, Symmetry | 13/17 | 24.6 ± 3.2 | Class I, Class III | ||||
Thiesen (2018) | Retrospective | 120 | 41/79 | 30.58 ± 9.46/ 19–57 | Caucasian | Class I | |
40, Relative asymmetry | 10/30 | 31.10 ± 9.89/ 19–51 | Gn < 2 mm | ||||
40, Moderate asymmetry | 15/25 | 30.57 ± 9.32/ 19–51 | Gn = 2–4 mm | ||||
40, Severe asymmetry | 16/24 | 30.05 ± 9.35/19–57 | Gn > 4 mm | ||||
Udomlarptham (2018) | Retrospective | 37 | 15/22 | 25.76 ± 7.14 | Asian | Class III | Me > 4 mm |
Wong (2018) | Retrospective | 59 | n/r | n/r | Asian | n/r | Me > 4 mm / 4° |
Zhang (2018) | Retrospective | 12 6, Asymmetry 6, Symmetry | n/r | n/r | Asian | n/r | Me and Pog > 4 mm |
Zheng (2018) | Experimental | 30 | 14 / 16 | 18–34 | Asian | n/r | Soft tissue chin > 1 mm |
Choi (2019) | Retrospective | 40 4, Class I 3, Class II 33, Class III | 18/22 | 25.5, 19–42 | Asian | Class I, Class II, Class III | n/r |
Kwon (2019) | Retrospective | 46 | 27/19 | 22 ± 4.8 | Asian | Class III | Me > 4 mm |
Tan (2019) | Computational | 10 | n/r | 20–75 | Asian | n/r | n/r |
Vernucci (2019) | Retrospective | 15 7, Symmetry 8, Asymmetry | 6/9 | 16–52 | Caucasian | Class I, Class III, Condylar hyperplasia, Hemifacial microsomia | n/r |
Han (2020) | Experimental | 29 | 15 / 14 | 23.1 ± 6.9 | Asian | Class III | Me > 4 mm |
Lee (2020) | Observational | 43 | 21 / 22 | 23.0 ± 8.20 | Asian | ||
10, Group 1 | 4 / 6 | 24.1 ± 10.14 | Class I | Me < 2 mm | |||
11, Group 2 | 5 / 6 | 22.3 ± 7.50 | Class III | Me < 2 mm | |||
9, Group 3 | 5/4 | 23.2 ± 5.19 | Class III | 2 mm < Me < 4 mm | |||
13, Group 4 | 7 / 6 | 22.6 ± 9.58 | Class III | Me ≥ 4 mm | |||
Ortún-Terrazas (2020) | Computational | 20 | 9 / 11 | M, 7.9 F, 8.2 | Caucasian | Unilateral crossbite | Minor < 0.3 mm; Moderate: malformations either in maxilla / mandible; Marked: maxillary and mandibular deformities + pronounced effect in the superficial soft tissue |
Zhu (2020) | Computational | 15 | n/r | n/r | Asian | n/r | Me > 3 mm |
Jo (2021) | Retrospective | 38 | Asian | n/r | Me > 4 mm | ||
23, PBO | 9/14 | 22.57 ± 4.97/ 17–37 | |||||
15, Grinding | 9/6 | 21.80 ± 4.89/ 18–38 | |||||
Lv (2021) | Prospective | 75 | 18–35 | Asian | Class I, Class II | ||
25, Class I symmetry | 6/19 | 23.46 ± 3.99 | Gn < 2 mm | ||||
25, Class II symmetry | 3/22 | 25.57 ± 4.55 | |||||
25. Class II asymmetry | 5/20 | 25.08 ± 3.59 | Gn > 4 mm | ||||
Mangal (2021) | Retrospective | 34 | 19/15 | 22.38 ± 5.20 /18–47 | Asian | Class III | Me > 4 mm |
Teng (2021) | Prospective | 122 | n/r | n/r | Asian | n/r | |
80, Asymmetry 42, Control | Chin > 2 mm | ||||||
Teng (2021) | Prospective | 40 | n/r | Asian | High angle Class III | ||
20, Experimental | 22.10 ± 3.01/ 18–28 | Me > 2 mm | |||||
20, Control | 24.10 ± 3.45/ 18–32 | ||||||
Ajmera (2022) | Retrospective | 42 | Asian | Class III | |||
21, Asymmetry | 7/14 | 23.0 ± 3.4 | Chin > 3 mm | ||||
21, Control | 7/14 | 23.0 ± 3.3 | |||||
Feng (2022) | Retrospective | 60 | n/r | Asian | n/r | ||
30, Symmetry | 26/ 20–32 | Me < 2 mm | |||||
30, Asymmetry | 24.7/ 19–30 | Me > 4 mm | |||||
Grissom (2022) | Ambispective | 54 | n/r | n/r | Caucasian | Goldenhar syndrome, Hemifacial microsomia, Mandibular hyperplasia, Mandibular hypoplasia, Unilateral condylar hyperplasia, Unilateral condylar destruction, Positional plagiocephaly, Juvenile arthritis | Chin > 4 mm |
Hsiao (2022) | Computational | 20 | 7/13 | 20–44 | Asian | Craniofacial dysplasia | n/r |
Ajmera (2023) | Retrospective | 42 | Asian | Class III | |||
21, Asymmetry | 7/14 | 23.0 ± 3.4 | Chin > 3 mm | ||||
21, Control | 7/14 | 23.0 ± 3.3 | n |
Author (year) | 3D Technique | Software used | MSP Type | Nomenclature | MSP construction | Reference points | Asymmetry assessment | Landmark digitization | Comparison | Reliability assessment | Measurements type | Outcome | Remark / Conclusion |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Wong (2005) | CT | n/r | Symmetry Plane | Best symmetry plane (BSP) | Mathematical optimization algorithm based | Landmark-independent | Symmetry value, representing the percentage of pixels that can be paired on both sides | Landmark-independent | BSP of midface vs BSP of mandible | n/r | Angular | 75.6% ↓ in asymmetry value using BSP | Correction of BSP of the mandible might greatly alleviate the FA |
BSP of midface and mandible diverse by 3.125° | |||||||||||||
Hartmann (2007) | FaceScan3D | Slim3D | Morphometric MSP | Symmetry plane | Iterative Closest Point (ICP) algorithm based OMR | Landmark-independent | Mean absolute distance between the original facial surface and mirror-image surface | Landmark-independent | 3D facial surface data recordings performed consecutively vs those performed on different days | Mean deviation angle between the symmetry planes | Linear, Angular | Among three measurement sets: phi φ, p < 0.05; dabs, p > 0.05 | The method introduced here can be used to determine the symmetry plane and degree of asymmetry using 3D facial data with convincing reproducibility and without having to refer to landmarks |
AlHadidi (2011) | CBCT | Insight SNAP | Cephalometric MSP | Plane passing through three median landmarks | Na-ANS-Ba | The 95th percentile surface distance measurement of each ROI | Manual | Mirroring using MSP vs mirroring using- registration based approach | Differences between repeated assessments of asymmetry | Linear | Mirroring using MSP = mirroring using registration, p > 0.05 | Both mirroring techniques provided similar quantification of mandibular asymmetry | |
Kim (2011) | CT | InVivoDental | Cephalometric MSP | Cranial MSP | Plane crossing three landmarks | CG-apFO-apFS | Absolute value of ⟂ distance from MSP to the midpoint | Manual | Cranial MSP vs Facial MSP | ICC, r ≥ 0.978 | Linear | Mean DCs 10–17 times > mean DFs, p < 0.001 | The facial MSP was not in agreement with the cranial MSP Cranial MSP could exaggerate the result of the jaw deviation |
Facial MSP | Plane crossing CG and vertically bisecting a line formed by FZS on both sides | CG-FZS | |||||||||||
Baek (2012) | 3D-CT | Simplant Pro | Cephalometric MSP | MSP | Plane passing through Na, Cl and Ba | Na-Cl-Ba | Distance of 3D landmarks to MSP | Manual | Four groups with distinct facial asymmetry features | ICC, 0.81 – 0.96 | Linear | Group 1 ≠ Group 2 ≠ Group 3 ≠ Group 4, p < 0.05 | Patients with asymmetry were classified into four statistically distinct groups according to their anatomic features |
Angular | Group 1 = 44%, caused by lateralization of mandibular body | ||||||||||||
Damstra (2012) | CBCT | SimPlant®Ortho Pro 2.1 | Cephalometric MSP | Plane passing through three midline structures | 1: S–N-ANS 2: S–N-Me 3: LFM-ACP-N 4: S–N 5: Cg-ACP 6: ELSA-MDFM | Linear measurements (mean ± SD) compared between AG and SG | Manual + Digital | Morphometric MSP vs Cephalometric MSPs | Method error (mean = 0.39 mm; 95% CI = 0.31–0.47 mm) | Linear | 1–3 and 6 vs Morphometric MSP, Absolute error (AE) > 1.00 mm | A morphometric approach to determine the MSP, might be more valuable for diagnosis and treatment planning of craniofacial asymmetry | |
Plane passing through two midline structures + ⟂ to HRP | |||||||||||||
r = 0.845–0.999 | 4–5 vs Morphometric MSP, AE < 1.00 mm | ||||||||||||
Morphometric MSP | Procrustes analysis (PA) based OMR | SOF, MZF, FNM, FOM | |||||||||||
Berssenbrügge (2014) | 3-camera fringe projection system | n/r | Morphometric MSP | Symmetry plane | ICP algorithm based OMR | Landmark-independent | 3D asymmetry index (AI) | Landmark-independent | 2D vs 3D methods | n/r | Linear | 3D AI, 2.54 ± 0.718 | An overall symmetry plane does not necessarily have to pass through facial midline points. This technique is assumed to give a better estimate of the facial symmetry plane than those that are based on only a few reference points or even solely on facial midline points |
2D AI | 2D AI, 24.8 ± 7.02 2D AI vs. 3D AI, r = 0.294; p = 0.038 | ||||||||||||
2D z-score | 2D z-score vs. 3D AI, r = 0.567 p < 0.001 | ||||||||||||
2D FA | 2D FA vs. 3D AI, r = 0.104; p = 0.472 | ||||||||||||
Wong (2014) | CT | Self-developed imaging-processing software | Symmetry plane | Optimal Symmetry plane (OSP) | Mathematical optimization algorithm based | Landmark-independent | n/r | Landmark-independent | Traditional surgical plan compared with Matching OSP based surgical plan | n/r | Significant ↓ in mandibular deviation | The new method resulted in surgical plans that brought about significantly less postoperative mandibular deviation while maintaining a reasonable occlusion | |
Linear | ADD, p = 0.046 PDD, p = 0.007 | ||||||||||||
Angular | DA, p = 0.001 | ||||||||||||
Gateno (2015) | n/r | n/r | Morphometric MSP | Primal sagittal plane (PSP) | LAGER (Landmark Geometric Routine) algorithm based on PA | All facial landmarks | n/r | n/r | n/r | n/r | n/r | n/r | Primal sagittal plane should improve the correctness of our cephalometric measurements and surgical plans |
Kim (2015) | CT | Vworks + Vsurgery | Cephalometric MSP | Midsagittal reference plane (MRP) | M1: Plane passing through two midfacial landmarks and ⟂to HRP and CRP | CG-P | Difference between x-coordinate (Δx), severity of asymmetry (SA), and direction of deviation (Dd) for four landmarks | Digital | M1 vs M2 | n/r | Linear | Δx, M1 ≠ M2, p < 0.05 SA, M1 ≠ M2 Dd, M1 ≠ M2 for Na and L1 Location of landmark, M1 ≠ M2, p < 0.05 | Location of midfacial landmarks, distance and direction of deviation, as well as the severity of asymmetry, may be influenced by the method of establishing the MRP |
M2: Plane passing through three midfacial landmarks | Op-CG-ANS | ||||||||||||
Ryu (2015) | CBCT | OnDemand3D | Cephalometric MSP | MSP | Plane constructed with N and ⟂ to the line connecting bilateral frontozygomatic point | N-FZP | Angular and Linear distance of the landmarks from the MSP | Manual | Asymmetry vs Control group | ICC, 0.82 – 0.93 | Linear | Shift, Yaw: Asymmetry ≠ Control, p < 0.05 | Me deviation in skeletal Class III deformity with mandibular asymmetry is influenced by rotation of mandibular posterior dentofacial structures |
Angular | Shift, Yaw: Hyperdivergent > Control; Hypodivergent > Control, p < 0.01 | ||||||||||||
Lee (2016) | 3D-CT | V-works | Cephalometric MSP | MRP | Plane passing through Op, Cg and ANS | Op-Cg-ANS | Distance of Me from the MRP | Manual | L ocation of the Me determined by PA cephalogram and 3DCT | n/r | Linear | Me deviation, PA cephalogram ≠ 3DCT Δx = 2.45 ± 2.03 mm, p < 0.05 | In facial asymmetry analysis using 3D CT, the definition of facial asymmetry should be based on Me deviation on 3D CT, not on the cephalogram |
Shin (2016)* | CBCT | Ondemand3D | Morphometric MSP | Symmetric MRP | PA based OMR | FO, FR, FS, FZM, GPC, HGC, Io,Or, Po, So, ZMS, A-point, ANS, Ba, INC,Na, NPC, Op, PNS, S, B-point, Me, G* | By comparing 1aandmark changes and differences in the amount of asymmetry between the original and symmetric configurations and 2 shapes produced by the 2 methods | Digital | Symmetric | ICC | Linear | Asymmetries measured by 3-landmark MRP > symmetric MRP | Statistical shape analysis confirmed that 3D-MRP constructed of Na, ANS, and PNS is compatible with the symmetric MRP and could be a valuable tool for evaluation of patients with FA |
MRP vs 3-landmark-based MRP | Intraexaminer, 0.924–0.941 | ||||||||||||
Interexaminer, 0.811–0.916 | SSED, SPD, 3-landmark MRP ≈ Symmetric MRP; p > 0.05 | ||||||||||||
Cephalometric MSP | 3-landmark-based MRP | Plane passing through three midline landmarks | Na-ANS-PNS | Total variations of asymmetry measurements according to 2 methods | |||||||||
Song (2016) | CT | Invivo 5.4 | Cephalometric MSP | MSP | Plane ⟂to FH plane and passing through N and S | N-S | Distance of Me from the MSP | Manual | CS vs POGS | ICC > 0.99 | Linear | CS ≈ POGS, p > 0.05 | POGS may be a clinically acceptable alternative to CS |
Sangln An (2017) | 3D CT | Simplant version 14.0 | Cephalometric MSP | Plane passing through three landmarks while ⟂ to FH plane | 1. FH-Na-Ba 2. FH-Na-S 3. FH-Cg-Ba 4. FH-Cg-S 5. Ba-Na-S 6. Ba-Cg-S 7. Ba-Na-ANS 8. Ba-Cg-ANS | Absolute values of differences in the measurement of Me deviation, ANS deviation, A-P line deviation | Manual | Eight different MSP configurations were compared | ICC, 0.91 – 0.95 | MSP 1, showed smallest absolute values for | Using MSPs passing through 3 median landmarks in the cranial base can lead to underestimation of the asymmetry of Me, ANS, and the A-P line. The authors suggest using MSPs perpendicular to the FH plane or a plane passing through ANS in clinical practice | ||
Linear | AVDMe – 0.81 ± 1.33 AVDANS – 0.44 ± 0.66 | ||||||||||||
Angular | AVDAP – 0.43 ± 0.59 | ||||||||||||
Dobai (2018)# | CBCT | CranioViewer software | Cephalometric MSP | Regression planes | Fifty planes were generated by a combination of unpaired landmarks, and paired cephalometric points | Combination of three unpaired landmarks and three paired cephalometric points | n/r | Digital | Regression planes vs Na-ANS-PNS (reference) plane | ICC > 0.9 | Angular | Regression planes generated from unpaired landmarks and paired points had < 5° deviation from the reference plane | The N-ANS-PNS reference plane, which represents the ideal morphometric midplane, can be substituted by planes derived from the following landmark combinations: ANS-G-Ba, ANS-G-S, ANS-S-De, PNS-G-Ba, PNS-S-Ba, and PNS-ANS-G, and PNS-N-Ba |
Economou (2018) | CBCT | Mimics | Cephalometric MSP | MRP | Plane ⟂ to the axial and coronal planes and passing through N | N | Distance of landmarks from the MRP | Manual | Hard tissue vs Soft tissue asymmetry | ICC, 0.74 – 0.98 | Linear Angular | Pog showed largest deviation from MRP | Soft tissue pogonion and gonion were identified as the most appropriate landmarks to clinically predict hard tissue facial asymmetry. Facial asymmetries are most pronounced in the lower facial third in patients with juvenile idiopathic arthritis |
Jajoo (2018) | 3D CT | 3D Studio Max | Morphometric MSP | PSP | LAGER algorithm based on PA | 11 unpaired and 26 paired landmarks | n/r | Digital | Algorithm-generated MSP vs Ground truth | n/r | Linear Angular | For all the algorithm-generated MSPs, DistN < 1 mm, DisU1 < 1 mm, DistPg < 2 mm and θ < 2° | All the LAGER algorithm-generated MSPs qualified as clinically acceptable. LAGER algorithm can be used clinically to determine the MSP for patients with CMF deformities |
Oh (2018) | CT | V-works | Cephalometric MSP | MRP | Plane ⟂to FH plane and passing through Cg and Op | Cg-Op | Linear and angular position of the condyle from MRP | Manual | Asymmetry vs symmetry groups | ICC, 0.98 – 0.99 | Linear | Mediolateral condylar position, p > 0.05 | In individuals with facial asymmetry, menton deviation is associated with the right/left differences caused by a smaller condyle on the deviated side |
Angular | Condylar angle, p > 0.05 | ||||||||||||
Volumetric | p < 0.05 | ||||||||||||
Thiesen (2018) | CBCT | SimPlant Ortho Pro | Cephalometric MSP | MSP | Plane passing through N and Ba and ⟂to FH plane | N-Ba | Distance of landmarks from the MSP | Manual | Relative vs Moderate vs Severe asymmetry | ICC > 0.80 | Linear Angular | Severe asymmetry: Contl side ≠ Dev side, p < 0.05 | A great deviation of the mandibular dental midline may indicate severe skeletal asymmetry in Class I adults |
Udomlarptham (2018) | CBCT | Simplant O & O | Cephalometric MSP | MSP | Plane ⟂to FH plane and passing through N and Ba | N-Ba | Distance of 3D landmarks to MSP | Manual | 2DP vs 3DS | Measurement error: | 2DP ≠ 3DS, p < 0.05 | The deviated centre landmarks to the MSP improved significantly, and improved surgical outcomes were achieved through 3DS | |
Linear, 0.43 – 0.92 | Linear | Go to MSP, p < 0.05 | |||||||||||
Angular, 0.39º—0.85º | Angular | Yaw angle, p < 0.05 | |||||||||||
Wong (2018) | CT | Self-developed imaging-processing software | Symmetry plane | OSP | Mathematical optimization algorithm based | Landmark-independent | Deviation angle and deviation distance formed by 2 OSPs in 3-dimensions | Landmark-independent | n/r | ICC, 0.99 | Linear | ADD > PDD, p < 0.0001 ADD, 7.22 ± 4.12 mm | Plane-to-plane analysis system (closely matching the mandibular OSP to the midface OSP) will correct misalignment and generally achieve a satisfactory overall skeletal symmetry |
Angular | FDA > I, p = 0.03 Mean FDA, 3.80° ± 3.89° 83% patients had significant mandibular misalignment (deviation, ≥ 4° or 4 mm) | ||||||||||||
Zhang (2018) | CBCT | Mimics | Morphometric MSP | Global registration | Preliminary MSP: N-S-Ba | Coordinate values of each landmark in 3D coordinate system | Digital | Interexaminer comparisons of the mean coordinate values of each landmark | ICC, > 0.9 | Linear | No significant difference in coordinate values by both examiners | The MSPs constructed using the novel method were extremely stable and reliable. The accuracy of MSPs does not rely on the accuracy of other planes and the MSP are not influenced by maxillofacial deformities, orbital malformations, or even mild or moderate cranial asymmetry | |
Final MSP: landmark-independent | |||||||||||||
Zheng (2018) | CT | ProPlan CMF® | Cephalometric MSP | Orbital margin plane (OMP) | Plane passing through the midpoint of the NFS and ⟂ to FZ suture line | NFS | Distance of the landmarks to MSP | Digital | OMP vs SBP | Paired t-test, p = 0.873 | Linear | Measurements in OMP < SBP, p < 0.05 | OMP is more stable, accurate, and reliable, and therefore more suitable for the evaluation of FA |
Skull base plane (SBP) | Plane passing through three landmarks | S–N-Ba | |||||||||||
Choi (2019) | 3D-CT | Mimics | Cephalometric MSP | MSP | Plane ⟂to AxP and passing through Cr and Cl | Cr-Cl | Distance of Me from the MSP | Manual | n/r | ICC, 0.91 – 0.99 | Linear | Chin deviation correlated with mandibular length (r = -0.897) and mandibular body length (r = -0.318) | Treatment planning in patients with chin deviation should involve a careful evaluation of the asymmetry of the upper and middle facial thirds |
Kwon (2019) | CBCT | Invivo 6 | Cephalometric MSP | MSP | Plane ⟂to FH plane and passing through N and S | N-S | Similarity Index (SI) and Non-overlapping volume (NOV) | Digital | Mandibular and lower facial soft tissue measurements between Dev and, N-Dev sides at T1 and T2 using MSP and AMP | ICC > 0.99 | Surface area, Volumetric | SI ↑ from 0.4 to 0.5, using MSP, and from 0.2 to 0.4, using AMP | SI and NOV can easily and intuitively evaluate overall 3D morphological discrepancies, especially 3D mandibular asymmetry |
Absolute mandibular midsagittal plane (AMP) | Plane passing through Me, B and G | Me-B-G* | |||||||||||
NOV, using MSP ≈ AMP | |||||||||||||
Tan (2019) | CT | Matlab | Symmetry plane | OSP | Oriented Bounding Box (OBB) → Mathematical translation + Mutual information method | Landmark-independent | n/r | Landmark-independent | Manual vs Semi-automatic | n/r | Linear | FAI, Manual ≈ Semi-automatic | Accuracy of semi-automatic method is almost equal to the accuracy of the doctor’s manual method |
Vernucci (2019) | CBCT | Dolphin | Cephalometric MSP | Anatomical MSP | Plane passing through Na, PCM and Ba | Na-PCM-Ba | Distance of 3D landmarks to MSP | Manual | Anatomical MSP vs Median plane | n/r | Linear | Anatomical MSP accuracy > Median plane | Anatomical MSP can be used as a reliable reference plane for transverse measurements in 3D cephalometry in cases of symmetrical or asymmetrical malocclusion |
Median plane | Plane passing through the midpoint of inter-zygomatic distance | Zr-Zl | Inter-zygomatic distance on PA cephalograms | AMD ≈ 1 mm; Percentage difference < 3% | |||||||||
Han (2020)* | CBCT | Invivo 6 | Cephalometric MSP | Facial MSP | Plane passing through the landmarks while ⟂ to FH plane | N-S | SI using mirroring | Digital | Facial MSP vs modified MSP configurations | ICC > 0.99 | SI using cmAMP > other MSPs, p < 0.05 | The cmAMP plane best matches the two anterior segments of hemi-mandible symmetrically and is closest to Facial MSP after orthognathic surgery in skeletal Class III patients with FA | |
AMP | Plane passing through Me, B and G | Me-B-G* | SI using cmAMP = SI using Facial MSP, p > 0.05 | ||||||||||
mAMP | Plane passing through the center point of Mf and ⟂ to the line connecting bilateral Mf | Mf | Linear | ||||||||||
cmAMP | Plane was established at a point with highest SI and at the centre of bilateral Mf | Mf | Angular | The distance (1.15 ± 0.74 mm) and angle (2.02 ± 0.82◦), between Facial MSP and cmAMP < between Facial MSP and other MSPs, p < 0.05 | |||||||||
Lee (2020) | CBCT | Invivo 5.4 | Cephalometric MSP | Plane passing through median landmarks and ⟂ to FH plane passing through bilateral landmarks | Reorientation method (RM)1: Cg-Ba and Ror-Rpo-Lor | ⟂distances from each landmark to three different MSPs | Digital | Three MSPs established by different RMs were compared | ICC > 0.9 | Linear | Mean absolute difference (MADs), RM 1 ≈ RM 2 ≈ RM 3, p < 0.05 | Although the differences in distance among the three MSPs were minor, the MSP established by RM 1 best approximated the true symmetrical MSP. This MSP could be implemented as the reference plane for the diagnosis of FA regardless of the extent of chin deviation | |
RM 2: N- IF-Ba and Ror -Rpo | MAD scores of RM 2 and RM 3 were 2–3 times > RM 1 (0.20 ± 0.10 mm) | ||||||||||||
RM 3: N, ANS, PNS and Ror -Rpo | |||||||||||||
Ortún-Terrazas (2020)† | CBCT | i-CAT | Morphometric MSP | Sagittal midplane | Principal Component Analysis (PCA) + ICP algorithm based OMR | Me, PhT, and G† | Distance from the midplane of the mandible (ManDev) and the distance from the midplane of the Me (MeS), to the sagittal midplane respectively | Manual | n/r | n/r | Linear | Bilateral measurements of cross side ≠ non-cross side, p < 0.05 | ManDev was more representative of the asymmetry than the MeS. PCA-based algorithm identified accurately and objectively the sagittal midplane in each subject, allowing the subsequent 3D-diagnosis workflow |
Significant malformations in mandibular ramus length (0.0086), maxillary palate width (0.0481), condylar head width (0.0408) in patients with severe asymmetry (jaw deviation > 0.8 mm) | |||||||||||||
Zhu (2020) | Face Scan 3D | Geomagic Studio 2013 | Morphometric MSP | PA Symmetry Reference Plane (SRP) | PA based OMR | Thirty-two anatomical landmarks | n/r | Manual | PA SRP vs WPA SRP vs Ground truth SRP | n/r | Linear | Global and regional position errors, WPA SRP < PA SRP | This novel automatic algorithm, based on weighted anatomic landmarks, can provide a more adaptable SRP than the standard PA algorithm when applied to severe mandibular deviation patients and can better simulate the diagnosis strategies of clinical experts |
WPA SRP | Weighted PA (WPA) based OMR | ||||||||||||
Ground truth SRP | Professional (regional ICP) algorithm based OMR | Angular | FAI error and Angle error, WPA SRP ≈ Ground truth SRP | ||||||||||
Jo (2021) | CBCT | OnDemand3D | Cephalometric MSP | MSP | Plane ⟂to FH plane and passing through N and S | N-S | Distance of landmarks from the MSP | Manual | PBO vs GR | ICC, 0.82 – 0.92 | Linear | PBO ≠ GR, p < 0.014 | PBO is recommended over the grinding method for patients with severe facial asymmetry |
Lv (2021 | CBCT | Dolphin 3D | Cephalometric MSP | MSP | Plane ⟂ to the horizontal plane and passing through N and Ba | N-Ba | Distance of landmarks from the MSP | Manual | Asymmetry vs Symmetry groups | ICC > 0.95 | Linear | Dev ≠ Contl side Co-MSP, p = 0.030 Go-MSP, p = 0.003 | Patients from the Class II asymmetry group showed significant differences between measurements on the contralateral and deviated sides, |
Angular | ∠C-MSPº: p = 0.022 | ||||||||||||
Mangal (2021) | CBCT | Invivo 6 | Cephalometric MSP | cmAMP | Plane ⟂to FH plane and passing through N and S | N-S | SI and NOV | Digital | SI and NOV between each segment and total mandible at T1 and T2 | Paired t-test, p < 0.001 | n/r | T2 – T1: SI score, between total mandible and anterior (r = 0.34, p = 0.044) and middle (r = 0.85, p < 0.001) segments | cmAMP based ToSS protocol allows accurate identification of the region of deformity of the mandible and minimizes residual asymmetries |
Teng (2021) | CBCT | Mimics | Cephalometric MSP | MSP | Plane passing through S, N and Anterior Nasal Spine | S–N-ANS | Deviation between mental apex of the chin and midsagittal plane in the coronal position | Manual | Jaw deformity vs Control | n/r | Linear | Jaw deformity ≠ Control, p < 0.001 | A positive correlation was found between the inclination of the occlusal plane and the degree of jaw deformity, with a linear relationship between them |
Angular | Positive correlation between MSP and Occlusal plane, Mental apex of chin Max. and Mand. Incisor midline ≠ MSP, p < 0.001 | ||||||||||||
Teng (2021) | CBCT | Mimics | Cephalometric MSP | MSP | Plane passing through S, N and Anterior Nasal Spine | S–N-ANS | Deviation between mental apex of the chin and midsagittal plane in the coronal position | Manual | Experimental vs Control | ICC, 0.97–0.99 | Linear | Experimental ≠ Control, p < 0.05 | Certain characteristics of mandibular symmetry and the occlusal plane were found in patients with high-angle skeletal class III malocclusion and jaw asymmetry |
Angular | Positive correlation between Mandibular deviation and Occlusal plane, r = 0.860, p < 0.001 | ||||||||||||
Ajmera (2022) | CBCT | 3D Slicer | Cephalometric MSP | MSP | Plane ⟂ to HP and passing through N and S | N-S | Distance of landmarks from the MSP | Manual | Asymmetry vs Control | ICC, 0.90–0.99 | Linear | T2, Significant correction of Me deviation, p < 0.001 | Despite significant correction after bimaxillary surgery, asymmetry persisted at several sites, thereby requiring secondary correction |
Residual asymmetry at MF, p < 0.001 | |||||||||||||
Feng (2022) | CBCT | Mimics | Morphometric MSP | MSPACB | Global registration | Euclidean distance of midline points to the MSP | Digital | MSPACB compared with MSPmorph | ICC = 0.99 | Linear | MSPACB ≈ MSPmorph, p > 0.05 | MSPACB is reliable for patients with or without facial asymmetry in maxillofacial asymmetry analysis | |
MSPmorph | PA | SOF, MZF, FNM, FOM | Stability, MSPACB > MSPmorph, p < 0.05 | ||||||||||
Grissom (2022) | CT/CBCT | Anatomic Aligner | Cephalometric MSP | MSP | Plane ⟂to axial plane and passing through Na and Ba | Na-Ba | n/r | Manual | Axial-plane-first vs midsagittal-plane-first | ICC > 0.89 | n/r | Facial reference frames defined by the midsagittal plane-first method ≠ axial-plane-first method, p = 0.001 | Midsagittal plane-first sequence improves the facial reference frames compared with the traditional axial-plane-first approach |
Morphometric MSP | PSP | Iterative WPA | Landmark-independent | ||||||||||
Hsiao (2022) | CT | n/r | Cephalometric MSP | LSP | Plane passing through CG, ANS and mid point of OrR_OrL | CG-ANS-mid OrR-OrL | Hausdorff distance (HD), Jaccard similarity coefficient (JSC) and Dice similarity coefficient (DSC) | Manual | Landmark-based vs Surface-based vs Voxel-based | n/r | Linear | HD: OSP < LSP < SSP JSC and DSC:OSP > LSP = SSP | The voxel-based method proposed in this research is a robust and reliable approach to evaluate the symmetry plane for severe asymmetry cases |
Morphometric MSP | SSP | PCA + ICP algorithm | Landmark-independent | Landmark-independent | |||||||||
OSP | Voxel-based symmetry plane | Landmark-independent | Landmark-independent | ||||||||||
Ajmera (2023) | CBCT | 3D Slicer | Cephalometric MSP | MSP | Plane ⟂ to HP and passing through N and S | N-S | Asymmetry index (AI) | Manual | Asymmetry index vs Asymmetry scores | ICC, 0.90–0.99 | Linear | MPA≈CDM > PA | Modified Procrustes analysis is proficient in evaluating cranio-facial asymmetry with more valid clinical representation and has potential applications in assessing asymmetry in a wide spectrum of patients |
MATLAB | Morphometric MSP | Clinically derived midline | N-S | Asymmetry scores | Digital | ||||||||
PA | All landmarks | ||||||||||||
Modified PA | Por-Or |
Study quality assessment
Certainty assessment | Impact | Certainty | Importance | ||||||
---|---|---|---|---|---|---|---|---|---|
№ of studies | Study design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | |||
Effectiveness of the technique (assessed with: Reliability assessment, Study outcomes, etc.) | |||||||||
42 | observational studies | not serious | seriousa | not serious | seriousb, c | strong association all plausible residual confounding would reduce the demonstrated effect | Cephalometric MSP and landmark dependent morphometric MSP are equally effective methods for MSP construction | ⨁⨁◯◯ Low | CRITICAL |
Ease of clinical applicability (assessed with: Technique simplicity, automation, etc.) | |||||||||
42 | observational studies | not serious | seriousd | not serious | not serious | very strong association | A fully automated MSP construction method may be more practical for clinical application | ⨁⨁⨁◯ Moderate | CRITICAL |