The aim of the study is to evaluate the reliability of landmark identification between manual & digital landmark plotting in both X and Y axis
MATERIALS AND METHODOLOGY:
50 pre-treatment lateral cephalograms were selected from patients reported for orthodontic treatment. The digital images of each cephalogram were imported directly into Dolphin software for onscreen digitalisation; while for manual tracing images were printed using a compatible X–ray printer. After the images were standardised and 10 commonly used hard tissue landmarks were plotted on each cephalogram by 6 different professional observers and the values obtained were plotted in X and Y axis.ICC was used to determine the intra-rater reliability for repeated landmark plotting obtained by both the methods.
The inter class correlation for manual in X and Y axis had a high reliability for all the 10 hard tissue points but when the intra class correlation was performed, all the hard tissue landmarks showed high reliability both in X and Y axis except POINT B which had moderate reliability with less agreement for cephalometric variables in X axis.
The inter and intraclass correlation in X and Y axis shows high reliability in hard tissue.
KEYWORDS: Cephalometric, dolphin software, landmark plotting,intraclass correlation
In Orthodontics, Cephalometrics plays a crucial role in diagnosis and treatment planning. Cephalometric radiograph is the product of a two- dimensional image of the skull in lateral view, which helps in enabling the relationship between teeth, bone, soft tissue, and empty spaces in horizontal and vertical planes of space. It also helps in evaluation, diagnosis, treatment results and prediction of growth.
The era of radiographic Cephalometry began in Orthodontics in 1931 by Broadbent and Hofrath; happen to simultaneously present a standardized cephalometric technique for obtaining standardized radiographs of the head(1)
The major sources of errors in cephalometric analysis are radiographic ﬁlm magniﬁcation, tracing, measuring, and landmark identiﬁcation. The inconsistency in landmark identification may lead to major error in cephalometric analysis.
“Dolphin” a cephalometric software, which according to manufacturer promises accurate landmark identification, cephalometric analysis, treatment prediction in orthognathic cases, superimposition, and acts as a tool for educating the patient on treatment outcome.
Several studies have been undertaken to compare the accuracy of landmark identification in scanned or digitised lateral cephalogram with the manual methods, whereas studies evaluating the reliability in landmark identification of digitally obtained radiographs with the manual method are scanty in the literature.
Hence, the objective of this study is to evaluate the errors and reliability in cephalometric landmark identification using Dolphin Orthodontic Software which is commercially available in the market for cephalometric analysis and to compare it with the manual cephalometric landmark identification.
AIMS AND OBJECTIVES
The aim of the study is to detect errors and reliability of landmark identification between manual & digital plotting for hard tissue landmarks.
The objective of the study is to conclude the superior method of plotting cephalometric landmarks & their reliability between manual and computerised method (DOLPHIN SOFTWARE) for hard tissue landmark points.
MATERIALS AND METHODOLOGY
Fifty pre-treatment digital lateral cephalograms of patients who reported to our department for orthodontic consultation and treatment were taken using digital cephalometer (Orthophos XG-SIRONA MODEL NO:D3352) and a written consent form was obtained from all the patients. The criteria for selecting the 50 cephalograms were
Good quality lateral cephalograms with sufficient contrast
Presence of permanent dentition with no missing and impacted teeth
Patients without trauma, syndromes, craniofacial deformity or gross asymmetry.
Lateral cephalograms of patients between 18-25 years of age
The originally saved digital cephalographic images are retrieved from the computer in which they were stored. As suggested by Alexander(2) among the selected images 3 registration crosses for orientation were marked, 2 in cranium and 1 in cervical vertebrae region for reorientation and 2 fiduciary points were chosen on the rulers that were imaged with the patients. Y axis was constructed by the software connecting the two fiduciary points as the vertical reference for landmark co-ordination, the X axis was constructed perpendicular to this line that served as horizontal reference(3) and they were printed to 100% of the original size. The prints were obtained in Fuji Medical Dry Imaging Film of size 20x25 cms (8”x10” inches)(4)
These printed lateral cephalograms were subjected to manual plotting and the digital images cephalograms were imported to computer aided cephalometric software Dolphin imaging V.11.8 to perform the landmark plotting.
A total number of 10 landmarks (fig 1)were identified by 6 qualified investigators, each observer was to perform landmark plotting per each cephalogram 3 times manually and 3 times digitally, at a two week interval as performed by Sang-Hun Yu et al(5). These were the most commonly used landmarks which plays a significant role in routine orthodontic diagnosis and treatment planning. Not more than two radiographs were plotted at a given time to avoid errors due to operator fatigue in both the methods.
The 50 samples (5),(6) were manually plotted in a dark room over an x-ray view box(fig- 2)on acetate sheet of thickness 0.003” (6),(7),(8) with a 0.5 mm(8),(9) lead pencil by 6 qualified investigators. All plotted sheets by investigators were collected and each landmark, was separately transferred to graph sheet and the values of the landmarks were evaluated in X and Y axis.(8)
Digital plotting was performed by the same 6 investigators using Dolphin imaging V.11.8 software. The images were calibrated by dpi settings and viewed in a 15” LCD flat screen monitor (fig-3). The landmarks were manually identified using cursor controlled mouse (10). After plotting , each image were printed in 20x25 cms (8”x10” inches)and landmarks were transferred to the graph sheet to get values in X and Y axis(6).
A total of 12,000 values were obtained of which 6,000 X –component and 6,000 Y-component. It includes 600 reading for an individual landmark; the analysis was carried out using Statistical Package for Social Sciences version 16.0. Average values of each landmark plotting were calculated & presented as mean ± standard deviation.(TABLE 1) Differences in mean were analysed using analysis of variances. The level of significance was set at P<0.05. Intraclass Correlation Coefficient (ICC) is used to determine the intrarater reliability for both the techniques. According to Landis and Koch(11), the following ICC interpretation scale was used: poor to fair (below 0.4), moderate (0.41-0.60), excellent (0.61-0.80), and almost perfect (0.81-1).
To evaluate the reliability of landmark identification by using manual and computerised plotting, the results obtained for the hard tissue landmarks (TABLE 2) were
SELLA on evaluation in X axis, complete homogeneity was achieved between manual and computerised plotting. When compared both the techniques, ICC indicates good homogeneity. And also in Y axis the manual landmark plotting indicates high reliability and in digital plotting with an ICC value of 1.000 which denotes complete homogeneity. On direct comparison between both the methods, ICC denotes very high reliability of the cephalometric variable. Zwei-Chieng Chang et al(3) stated that sella is consistent and reliable which supports the values obtained for our study. According to Jia –Kuang Liu et al(12), their study stated that sella was the most accurately identified landmark. According to Y-J Chen et al(13) stated that the errors in sella were smaller than 1mm in both horizontal and vertical directions.Yi-Jane Chen et al(14), stated that the landmarks with minimal difference was sella and the values obtained were statistically significant. Scott R McClure et al(9) also described sella as the most accurately defined landmark in both film and digital methods.
NASION in X axis denotes complete homogeneity. On comparing both the technique, ICC is 0.863 which is good indicator for cephalometric reliability. On Y axis NASION intraclass correlation indicates excellent reliability in landmark plotting. Tancan Uysal et al(15) suggested that Nasion identification was difficult. Several other studies showed inconsistency in Nasion identification such as Baumrind S and Frantz et al(16), Scott R McClure et al(9) suggested that NASION point was considered accurate for both digital identification and film based landmark identification, which supports the results obtained from our study. YJ Chen et al(13) study stated that the landmarks with minimal location difference was the point NASION, compared to all the landmark points, he has studied. Zwei-Chieng Chang et al(3), in his study of landmark identification errors by CBCT image mode suggested that errors were rated for NASION, ORBITALE, ANS compared to other landmark identification points.
PORION in X axis ICC are close to homogeneity. When compared directly between manual and digital ICC value is 0.881, indicates good reliability. Along Y axis, ICC is 0.980 in manual and 0.999 in dolphin. When comparing both methods the ICC value is 0.912, which are closest to homogeneity. Eddie Hsiang –Hua –Lai et al(17) suggested that the difficulty in identifying PORION on images of superimposed structures results in variation of Frankfort horizontal plane. Lance Q.Brunpz et al(18) showed PORION had lower reliability in landmark identification, thus leading to significant unreliability of Frankfort horizontal plane. Yi Jane Chen et al(19) suggested that the reliability of PORION in digital images was inferior in our radiographs. Zwei chieng chang et al(3) had a difficulty in identifying PORION due to overlapping structures, thus increase in identification error. J Chen et al(13) had a very small error which is less than 1mm in both horizontal and vertical directions stating that PORION point was accurately plotted. This study supports our study for high identification reliability of PORION landmark. The other reason for identification of PORION may be due to conscious effort of the observers while plotting as the literature suggests otherwise.
ORBITALE - in X axis ICC values indicate complete homogeneity. And when both are directly ICC is 0.835 which indicates excellent reliability. When assessed along Y axis ICC value are reliable but when compared directly the ICC value is 0.671, indicating excellent agreement. Celik E et al(20), Sayinsu K et al(21), Chen Yi et al(19),Uysal T et al(22) and Bruntz L Q et al(18) suggested that ORBITALE sometimes is not clearly identified in a cephalogram.Yi Jane Chen et al(13), suggested that ORBITALE point had a maximum error difference compared to other landmarks he had studied in both X and Y axis. Zwei Chieng Chang et al(3) in his study he expressed those identification errors were greater for ORBITALE even with CBCT derived cephalograms. The explanation given for this later identifies errors was because of superimposed bilateral structures and maybe blurred images. Broch J, Slagsvold et al(23) suggested that the landmark if in blurred area of facial structures like ORBITALE errors will be larger. Chen Y J et al(13) suggested that significant location difference of PORION and ORBITALE can lead to alteration in FH plane.
ANS in X axis has an ICC value are reliable but when compared directly ICC value is 0.741, which shows excellent agreement. In Y axis ICC for ANS the digital plotting has a very higher value, it is very close to complete homogeneity. On comparing directly between manual and digital, ICC is 0.750 which indicates excellent agreement for cephalometric variable. Baumrind S et al(16), Santoro M et al (24), Polat –Ozsoy O et al(25), Ongkosuwito Em et al(26), Gregston MD et al(27) and Houston Wj et al(28) found difficulty in locating ANS. Yi Jane Chen et al(19) suggested that the reliability of ANS was inferior that of which original radiograph and hence it was difficult to compare absolute value of reliability between different studies.
ICC for PNS in X axis shows good agreement for manual plotting but for digital it is near to complete homogeneity but when both techniques are compared the ICC is 0.616 which shows good reliability for cephalometric variables. In Y axis intraclass correlation value is closest to complete homogeneity. Ralf Kurt Willy -Schulze et al(29), in his journal observed that PNS was least reliable landmark in the X and Y direction both for inter and intra observer reliability. Forsyth DB et al(30) suggested that PNS is a poorly defined structure that is disguised by surrounding noise. Lance Q Bruntz et al(18) suggested that PNS had low reliability in landmark identification as observed from inter observer error. S.S Huja et al(31) observed in his study indicating the value as more than 1mm,while taking upper 95 interval consideration suggesting that PNS identification was reliable during superimposition. Scott R McClure et al(9) suggested that PNS seems to be more reliably identified in vertical than in horizontal directions.
POINT A on digital plotting along X axis has a higher value compared to manual method; significant difference between both the values is not present. But on direct comparison, the ICC value is 0.771 which suggests good reliability as a cephalometric variable. In Y axis digital plotting has a higher value than in manual method but both are near to homogeneity. And in comparison of both manual and digital, the ICC value is 0.749 indicating excellent reliability as the value is within 0.61 and 0.80. Scott R McClure et al(9), in his study indicated that POINT A was accurate in horizontal direction. Trpkova et al(29)found POINT A to be accurate vertically.Sohrab Shaheed et al(32) had a problem in accurately identifying the A POINT. Jacobson R L et al(33) in his landmark article where he revisited POINT A, suggested this point is obscured by prominent cheeks, and rare earth identifying screens for enhancement of soft tissue visualisation which makes it difficult to locate accurately. Serge kazandjian et al(34), suggested that intra operator reliability was decreased for POINT A, in Y axis. Priscila De Araujo Guedes et al(35), had a difficulty in locating POINT B in both manual and computer assisted methods.
POINT B, with ICC values in X axis are near to complete homogeneity. On direct comparison, the ICC is 0.737 in x axis which denotes good reliability. For Y axis both the methods indicate complete homogeneity. On direct comparison between manual and digital methods intraclass correlation value is 0.469 which shows moderate agreement. Serge Kazhandjian et al(34), suggested that interoperator reliability was included by using computer assisted method for POINT B in both X and Y axis. Scott R McClure et al(9) suggested that identification of POINT B was accurate among the horizontal plane. Alok Rajendra Shah et al(7) suggested that POINT B, lies on a poorly defined outline or low contrast areas. Priscila De Araujo Guedes et al (35) had a difficulty in reproducing POINT B in both manual and computer assisted methods.
POGONION in X axis ICC values are in complete homogeneity. When compared directly the intraclass correlation value is 0.754, indicating good reliability. Along Y axis intraclass correlation values indicate complete homogeneity, as they range between 0.81 and 1. But when compared directly, the ICC is 0.665 which indicates good agreement.Neha Agarwal et al(8), have indicated location of POGONION was difficult. ZweiChieng-Chang et al(3), indicated that identification of POGONION in horizontal direction was relatively consistent and reliable in both imaging modes. The results of this study correlated with the results of our study.
MENTON when subjected to ICC in X axis both the values are near to complete homogeneity. On direct comparison, the ICC is 0.759 which is a good reliability of landmark plotting. Manual plotting in Y axis manual values shows good reliability and digital value indicates complete homogeneity, but when directly compared between both the techniques the ICC value is 0.718 indicating good reliability. Y –J Chen et al(13) suggested that the error for MENTON point were smaller than 1mm in both vertical and horizontal direction suggesting reliability. Serge Kazandjian et al(34) indicated that intraoperator reliability improved for MENTON point when compared to other points both in X and Y axis. The results of these studies support our findings. Yi Jane Chen et al(13) explained that the uncertainty in locating MENTON point may be caused by the difficult of delineating landmark on a curved anatomical boundary. According to Zwei Chieng Chang et al(3),landmark identification of MENTON in vertical direction was the most reliable point.
The interclass and intraclass correlation in X and Y axis shows high reliability in hard tissue.
The results obtained for manual and digital was almost similar but the digital landmark plotting has an added advantage in archiving, retrieval, transmission and can be enhanced during plotting of lateral cephalograms. So that the digital method of landmark plotting could be preferred for both daily use and research because of the advantages.
Broadbent B. H. A new x-ray technique and its application to orthodontia. The Angle Orthodontist 1931;1(2):45-66.
Chang Z-C,Hu F-C,Lai E, Yao C-C,Chen M-H,Chen Y-J. Landmark Identification Errors On Cone Beam Computed Tomography- Derived Cephalograms And Conventional Digital Cephalograms.American Journal Of Orthodontics And Dentofacial Orthopaedics.2011;140(6)
Ravi Kumar Mahto, Om Prakash Kharbanda, Ritu Duggal, Harish Kumar Sardana.A comparison of cephalometric measurements obtained from two computerized cephalometric softwares with manual tracings. 2016;50(3):162–70.JIOS
Sang H.Y. Reliability of landmark identification on monitor displayed lateral ceph image Am J Orthod Dentofacial Orthop 2008;133(6) 790-e1.
Paixao, Mariane Bastos, Marcio Costa Sobral, Carlos Jorge Vogel, and Telma Martins de Araujo. Comparative study between manual and digital cephalometric tracing using Dolphin Imaging software with lateral radiographs. Dental Press Journal of Orthodontics 2010;15(6):123-130.
Alok Rajendra Shah, Girish Karandikar, VK Ravindranath, Manish Sonawane AM. RESEARCH ARTICLE A Comparative Study of Reliability and Accuracy of Manual and Digital Lateral Cephalometric Tracing. J Contemp Dent. 2016;6(April):15–8.
Neha Agarwal, Dinesh K. Bhagga, Payal Sharma. A comparative study of the cephalometric measurements with digital versus manual methods. J Ind orthod soc 2011;45(2):85-90
McClure S.R., Sadowaky P.L., Ferreira A., Jacobson A. Reliability of digital versus conventional cephalometric radiology: a comparative evaluation of landmark identification error. Sem in Orthod 2005;11(2):98-110.
Houston W.J. The analysis of errors in orthodontic measurements. Am J Orthod Dentofacial Orthop 1983;83(5):382-390
Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74.
Liu J-K, Chen Y-T, Chenh K-S.Accuracy Of Computerised Automatic Identification Of Cephalometric Landmarks. American Journal Of Orthodontics And Dentofacial Orthopaedics.2000;118(5):535-40
Chen, Yi-Jane, Ssu-Kuang Chen, Jane Chung-Chen Yao, and Hsin-Fu Chang. The effects of differences in landmark identification on the cephalometric measurements in traditional versus digitized cephalometry. The Angle orthodontist 2004;74(2):155-161
Chen, Y. J., S. K. Chen, H. W. Huang, C. C. Yao, and H. F. Chang. Reliability of landmark identification in cephalometric radiography acquired by a storage phosphor imaging system. Dentomaxillofacial Radiology 2004;33(5):301-306.
Uysal, Tancan, Asli Baysal, and Ahmet Yagci. Evaluation of speed, repeatability, and reproducibility of digital radiography with manual versus computer-assisted cephalometric analyses. Eur J Orthod 2009;31(5):523-528.
Baumrind, S., & Frantz, R. C. The reliability of head film measurements: 1. Landmark identification. Am J Orthod 1971;60(2):111-127.
Lai, Eddie Hsiang-Hua, Curtis Ssu-Kuang Chen, Zwei-Chieng Chang, Chung-Chen Yao, and Yi-Jane Chen. Cephalometric analysis using digital radiography acquired by a storage phosphor imaging system: a comparison of reading soft-and hard-copies. Journal of Dental Sciences 2007;2(2):65-74.
Bruntz, Lance Q., J. Martin Palomo, Sally Baden, and Mark G. Hans.A comparison of scanned lateral cephalograms with corresponding original radiographs. Am J Orthod Dentofacial Orthop 2006;130(3):341.
Chen Y.J, Chen S.K, Chang H.F, Chen K.C. Comparison of landmark identification in traditional versus computer aided digital cephalometry. Angle Orthod 2000;70(5):387-392.
Celik E, Polat-Ozsoy O, Toygar Memikoglu TU. Comparison of cephalometric measurements with digital versus conventional cephalometric analysis. Eur J Orthod. 2009;31:241–246.
Sayinsu, Korkmaz, Fulya Isik, Göksu Trakyali, and Tülin Arun. An evaluation of the errors in cephalometricmeasurements on scanned cephalometric images and conventional tracings. Eur J Orthod 2007;29(1):105-108.
Uysal, Tancan, Asli Baysal, and Ahmet Yagci. Evaluation of speed, repeatability, and reproducibility of digital radiography with manual versus computer-assisted cephalometric analyses. Eur J Orthod 2009;31(5):523-528.
J Broch,O Slagsvold,M Rosler,Errors In Landmark Identification In Lateral Radiographic Headplates. Eur J Orthod (1981)3;(1):9-13
Santoro M, Jarjoura.K, Cangialosi.T. Accuracy of digital and analogue cephalometric measurements assessed with the sandwich technique. Am J Orthod Dentofac Orthop 2006;129:345-51
Polat-Ozsoy O, Gokcelik A, Toygar Memikoglu TU. Differences in cephalometric measurements: a comparison of digital versus hand-tracing methods. Eur J Orthod 2009; 31: 254–259.
Ongkosuwito E.M., Katsaros C.,N Hof V.M.A, Bodegom J.C. The reproducibility of cephalometric measurements: a comparison of analogue and digital methods. Eur J Orthod 2002;24(6):655-665.
Gregston MD, Kula T, Hardman P, Glaros A, Kula K. A comparison of conventional and digital radiographic methods and cephalometric analysis software: I. hard tissue. Semin Orthod 2004; 10: 204–211.
Houston W.J, Maher R.E. McElroy. Sources of error in measurement from cephalometric radiographs Eur J orthod 1986;8(3):149-15.
Trpkova B, Major P, Prasad N, Nebbe B. Cephalometric landmarks identification and reproducibility: A meta analysis. Am J Orthod Dentofacial Orthop 1997;112:165-70.
Forsyth D. B, W. C. Shaw, S. Richmond, and C. T. Roberts. Digital imaging of cephalometric radiographs, part 2: image quality. The Angle Orthodontist 1996;66(1):43-50.
Huja, S. S., E. L. Grubaugh, A. M. Rummel, H. W. Fields, and F. M. Beck. Comparison of hand-traced and computer-based cephalometric superimpositions. The Angle Orthodontist 2009;79(3):428-435.
Sohrab Shaheed, Ayesha Iftikhar, Ghulam Rasool, Ulfat Bashir, Accuracy of linear cephalometric measurements with scanned lateral cephalograms. Pakistan Oral & Dental Journal 2011;31(1):68-72.
Jacobson RL, Jacobson A. Point A revisited. Am J Orthod 1980;77:92-96.
Kazandjian, Serge, Stavros Kiliaridis, and Anestis Mavropoulos. Validity and reliability of a new edge-based computerized method for identification of cephalometric landmarks. The Angle Orthodontist 2006;76(4):619-624.
Guedes PA, Souza, JEN de, Tuji FM, Nery EM A comparative study of manual vs. Computerized cephalometric analysis. Dental Press J. Orthod.2010.
FIGURE-1 3 Orientation marks-X and Y axis along the rulers.10 hard tissue landmarks.S-Sella,N-Nasion,Po-Porion,Or-Orbitale,ANS-anterior nasal spine,PNS-Posterior nasal spine,A-point A,B-point B,Pg-Pogonion,Mn-Menton.