396Angle Orthodontist, Vol 71, No 5, 2001

Original Article

Effect of Head Rotation on Lateral Cephalometric RadiographsYoung-Jooh Yoon, DDS, MSD, PhDa; Kwang-Soo Kim, DDS, MSD, PhDb; Mee-Sun Hwang, DDS,

MSDc; Heung-Joong Kim, DDS, MSD, PhDd; Eui-Hwan Choi, DDS, MSD, PhDe;Kwang-Won Kim, DDS, MSD, PhDf

Abstract: The purpose of this study was to identify the potential projection errors of lateral cephalo-metric radiographs due to head rotation in the vertical Z-axis. For this investigation, 17 human dry skullsamples with permanent dentition were collected from the Department of Anatomy in the College ofMedicine, Chosun University. They had no gross asymmetry and were well preserved. Each dry skull wasrotated from 08 to 6158 at 18 intervals. A vertical axis, the Z-axis, was used as a rotational axis to have527 lateral cephalometric radiographs exposed. The findings were that: (1) angular measurements havefewer projection errors than linear measurements; (2) the greater the number of landmarks on the midsag-ittal plane that are included in angular measurements, the fewer the projection errors occurring; (3) hori-zontal linear measurements decrease gradually in length as the rotational angle toward the film increases,whereas a small increase and then decrease of the length occurs as the rotational angle toward the focalspot increases; (4) horizontal linear measurements have more projection errors than vertical linear mea-surements according to head rotation; and (5) projection errors of vertical linear measurements increase asthe distance from the rotational axis increases. In summary, angular measurements of lateral cephalometricradiographs are more useful than linear measurements in minimizing the projection errors associated withhead rotation on a vertical axis. (Angle Orthod 2001;71:396–403.)

Key Words: Projection error; Head rotation; Lateral cephalometric radiograph

INTRODUCTION

Lateral cephalometric radiography has limited orthodon-tic applications in the precise interpretation of a patient’scraniofacial entity regardless of its standardization. Ce-phalographic projection is a central projection in which theobject is projected onto the film by a beam that divergesfrom the radiation source, the focal spot. Only in the rarelyoccurring ideal case is the projection surface where the filmis placed perpendicular to the central ray of the beam.1,2

Much research has been conducted into the reliability of

a Assistant Professor of Orthodontics, College of Dentistry, ChosunUniversity, Gwang-Ju, Korea.

b Private practice in Gwang-Ju, Korea.c Resident in Orthodontics, College of Dentistry, Chosun Univer-

sity, Gwang-Ju, Korea.d Assistant Professor of Oral Anatomy, College of Dentistry, Cho-

sun University, Gwang-Ju, Korea.e Associate Professor of Oral & Maxillofacial Radiology, College

of Dentistry, Chosun University, Gwang-Ju, Korea.f Professor of Orthodontics, College of Dentistry, Chosun Univer-

sity, Gwang-Ju, Korea.Corresponding author: Young-Jooh Yoon, DDS, MSD, PhD, De-

partment of Orthodontics, College of Dentistry, Chosun University,421, Seosuk-Dong, Dong-Ku, Gwang-Ju, 501-825, Korea(e-mail: yjyun@mail.chosun.ac.kr).

Accepted: February 2001. Submitted: December 2000.q 2001 by The EH Angle Education and Research Foundation, Inc.

the lateral cephalometric radiograph.3–16 Cephalometric er-rors are divided into 3 categories:4,5 (1) Identification errorsoccur in finding certain anatomic landmarks; (2) Projectionerrors are caused by the wrong positioning of the radio-graph; (3) Mechanical errors occur in drawing lines be-tween points on a tracing and in measuring the tracing witha ruler or protractor.

A head-holding device, consisting of an ear rod and nasalpositioner, is used for lateral cephalometric radiographs tominimize the projection errors caused by head rotation in thevertical, transverse, and anteroposterior axes. However, whenthe device is used to contact the external auditory meatus andthe soft tissue of the patient, the head can be incorrectly po-sitioned sagittally, anteroposteriorly, or vertically, because thepatient’s head can be slightly rotated within the head-holdingdevice. Because of such improper positions due to head ro-tation, an error can occur in cephalometric measurements. Be-cause of the errors caused by the different locations of thehead, cephalometric linear and angular measurements canvary depending on the different locations of anatomic struc-tures against the central ray.1,2 Therefore, unless the projectionerrors are precisely evaluated and understood, cephalometricmeasurements may have only limited application in orthodon-tics. The purpose of this study is to identify the potential pro-jection errors of lateral cephalometric radiographs due to headrotation in the vertical Z-axis.

397HEAD ROTATION AND LATERAL CEPHALOMETRIC RADIOGRAPHS

Angle Orthodontist, Vol 71, No 5, 2001

FIGURE 1. Positioning of the dry skull. 1. ear rod, 2. nasal positioner,3. rotational scale.

FIGURE 2. The rotational axes of skull.

MATERIALS AND METHODS

Materials

For this investigation, 17 human dry skulls with per-manent dentition were collected from the Department ofAnatomy in the College of Medicine, Chosun University.They had no gross asymmetry and were well preserved.Sex, occlusion, and skeletal pattern were not significantlyconsidered.

Methods

Lateral cephalometric radiographs. Before the radio-graph was taken, 8 anatomic landmarks were designated onthe human dry skulls. Steel balls of 1.0-mm diameter wereglued on each of 11 landmarks including bilateral land-marks. The FH (Frankfort horizontal) plane of the skull wasplaced parallel to the floor and was tightly positioned withan ear rod, headrest, and rubber bands (Figure 1). The anglewas marked on the upper part of the cephalostat, and anindicator was attached so that the angle could be read moreeasily. A PM 2002 PROLINE cephalostat machine (Plan-meca Co Ltd, Helsinki, Finland) was used for this investi-gation. The standard focus-median plane and film-medianplane distances were 135.5 cm and 13.5 cm, respectively.Each skull was rotated from 08 to 6158 at 18 intervals. Avertical axis, the Z axis, was designated as a rotational axisconnecting the center of both ear rods in the direction ofthe submentovertex, and 527 radiographs were taken basedon this axis (Figure 2). The code ‘‘1’’ means a rotationtoward the focal spot, and ‘‘2’’ means a rotation towardthe film.

Establishment of landmarks and measurements.

A. Cephalometric landmarks• S (Sella turcica): The midpoint of the sella turcica, de-

termined by inspection.

• N (Nasion): The intersection of the internasal sutureand the frontonasal suture in the midsagittal plane.

• A (Subspinale): The deepest point on the premaxillabetween the anterior nasal spine and prosthion.

• B (Supramentale): The deepest point in the concavitybetween the infradentale and the pogonion.

• Me (Menton): The most inferior point on the mandib-ular symphysis in the midsagittal plane.

• Corpus left: The left point of a tangent of the inferiorborder of the corpus.

• Ramus down: The lower point of a tangent of the pos-terior border of the ramus.

• Ar (Articulare): The intersection of the posterior mar-gin of mandibular condyle and temporal bone.

B. Cephalometric measurements1. Linear measurements

a. Horizontal linear measurements• Anterior cranial base length (S-N)• Mandibular body length (Gonion [Go]-Me)

b. Vertical linear measurements• Anterior facial height (N-Me)• Posterior facial height (S-Go)

2. Angular measurements• SNA• SNB• Saddle angle (N-S-Ar)• Articular angle (S-Ar-Go)• Gonial angle (Ar-Go-Me)• AB to mandibular plane angle

Input of data. The point at which the steel balls on thelandmarks including Corpus left, Ramus down, and Ar (Ar-ticulare) met the bone surface was marked on the film witha sharp metal pin. The results were input using a digitizerconnected with a Macintosh computer. Detailed measure-ments—with length units of 0.01 mm and angular units of0.018—were made using Quick Ceph Image Proy Version3.0 (Quick Cepht Systems Co, San Diego, California). Thesame person recorded the data throughout the whole pro-cess to reduce measurement errors.

Reliability of digitizing. Each result was digitized 3 times

398 YOON, KIM, HWANG, KIM, CHOI, KIM

Angle Orthodontist, Vol 71, No 5, 2001

TABLE 1. Comparison of the Measurements From Zero to Each Rotational Angle Toward the Filma

Measurement 2158 2148 2138 2128 2118 2108 298

Linear, mm

1. N-S

2. Go-Me

3. N-Me

4. S-Go

MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %

67.44***4.81

25.3772.67***5.96

25.78116.80***

8.8021.4985.669.39

20.23

67.74***4.67

24.9573.07***6.05

25.26116.91***

8.8221.4085.739.44

20.15

68.18***4.67

24.3373.566.07

24.63117.04***

8.7221.2985.769.47

20.11

68.44***4.69

23.9773.78***5.97

24.35117.19***

8.7921.6785.729.52

20.16

68.89***4.72

23.3474.15***6.05

23.86117.30***

8.8321.0785.889.610.02

69.23***4.66

22.8674.61***6.14

23.26117.44***

8.7520.9585.739.51

20.15

69.52***4.69

22.4675.15***5.99

22.57117.54***

8.7720.8785.789.49

20.08

Angular, 8

5. SNA

6. SNB

7. N-S-Ar

8. S-Ar-Go

9. Ar-Go-Me

10. AB/Go-Me

MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %

80.585.55

20.3481.016.85

20.35132.53

8.6220.51136.89**

8.971.10

116.396.13

20.1675.997.380.09

80.435.31

20.5181.026.78

20.34132.52

8.6120.52136.86**

9.201.08

116.406.23

20.1675.957.630.03

80.54*5.24

20.4081.056.79

20.43132.71

8.4020.37136.68**

8.660.95

116.326.09

20.2276.157.550.30

80.615.29

20.3181.116.83

20.24132.71

8.4620.37136.60**

8.630.88

116.215.97

20.3276.067.600.18

80.50*5.28

20.4481.086.84

20.27132.75

8.4020.34136.46*

8.710.78

116.386.05

20.1876.217.650.38

80.615.14

20.3181.086.75

20.28132.88

8.3620.25136.27*

8.840.64

116.516.08

20.0776.087.600.21

80.59*5.12

20.3281.156.86

20.19133.05

8.4420.12136.27*

8.680.64

116.22*6.14

20.3176.057.630.16

a Reference group: 00.* P , .05; ** P , .01; *** P , .001.

to analyze the digitizing errors and to produce an averagevalue for each angle. The coefficient constant (r) amongthe 3-time digitizing values was .9956, which may be con-sidered very reliable.

Statistical analysis. Four linear and 6 angular measure-ments were taken from 08 to 6158 at 18 intervals. Pairedt-tests were performed between the measurements from 08to each rotational angle using the SPSS (SPSS Inc, Chi-cago, Illinois) statistical program.

RESULTS

The changes in the cephalometric linear and angularmeasurements occurring with 08 to 2158 rotational anglesare presented in Table 1 and those from 08 to 1158 rota-tional angles in Table 2.

1. Linear Measurementsa. Horizontal linear measurements

• Anterior cranial base lengthThe anterior cranial base length was 71.27 mm at 08. Asthe rotational angle toward the film increased, its length

decreased, whereas its length increased and then decreasedas the rotational angle toward the focal spot increased. Thedegree of these changes was not significant. There was astatistical difference between the measurements from 08 toeach rotational angle toward the film (P , .05) and to ro-tations of more than 198 rotational angle toward the focalspot (P , .01). The difference was less than 1% from 08to 1128 and from 08 to 248 rotational angle. The maximumreduction was 25.37% at 2158 rotational angle.

• Mandibular body lengthThe mandibular body length was 77.13 mm at 08. As therotational angle toward the film increased, its length de-creased, whereas its length increased and then decreased asthe rotational angle toward the focal spot increased. Thedegree of these changes was not significant. There was astatistical difference between the measurements at morethan 228, except at 2138 rotational angle toward the film(P , .01) and more than 128, except from 1118 to 1138rotational angle toward the focal spot (P , .05). The dif-ference was less than 1% from 08 to each rotational angletoward the focal spot and from 08 to 248 rotational angle

399HEAD ROTATION AND LATERAL CEPHALOMETRIC RADIOGRAPHS

Angle Orthodontist, Vol 71, No 5, 2001

TABLE 1. Extended

288 278 268 258 248 238 228 218 08

69.72***4.70

22.1775.37***6.18

22.28117.71***

8.8120.7385.839.47

20.03

70.12***4.73

21.6175.74***6.14

21.79117.70***

8.8520.7385.879.430.02

70.29***4.72

21.3875.94***6.14

21.54117.81***

8.8020.6485.809.62

20.06

70.45***4.69

21.1576.18***6.26

21.23117.93***

8.8920.5485.909.510.05

70.62***4.74

20.9176.40***6.21

20.94118.12***

8.9220.3885.959.530.11

70.84***4.75

20.6076.64***6.25

20.64118.21***

8.7020.3085.959.580.11

70.95***4.75

20.4676.80**6.26

20.42118.29**

8.780.23

85.959.470.11

71.13*4.70

20.2076.816.19

20.41118.51

8.850.05

85.929.480.07

71.274.660.00

77.136.260.00

118.588.810.00

85.869.540.00

80.59*5.13

20.3281.156.78

20.19133.18

8.0720.02136.03

8.420.47

116.366.16

20.2076.117.610.23

80.625.12

20.2981.166.87

20.17133.17

8.2620.03135.86

8.660.34

116.426.18

20.1476.227.760.40

80.724.98

20.1781.256.78

20.06133.12

8.1720.06135.89

8.460.36

116.486.19

20.0976.097.670.21

80.775.09

20.1081.276.900.00

133.218.220.00

135.838.620.32

116.36*6.12

20.1976.097.720.23

80.674.90

20.2381.256.83

20.07133.31

8.210.07

135.768.630.27

116.41*6.01

20.1576.037.680.13

80.744.98

20.1581.286.90

20.03133.32

8.100.09

135.708.490.22

116.436.20

20.1376.027.820.13

80.835.04

20.0381.326.880.02

133.388.070.13

135.598.270.14

116.566.08

20.0275.997.710.09

80.755.03

20.1381.266.91

20.05133.45

8.180.18

135.578.190.13

116.525.98

20.0676.187.620.33

80.864.980.00

81.306.940.00

133.218.130.00

135.408.130.00

116.596.140.00

75.927.940.00

toward the film; the maximum reduction was 25.78% at2158 rotational angle.

b. Vertical linear measurements• Anterior facial height

The anterior facial height was 118.58 mm at 08. As therotational angle toward the film increased, its length de-creased, whereas its length increased and then decreased asthe rotational angle toward the focal spot increased. Thedegree of these changes was similar regardless of the di-rection of rotation. There was a statistical difference be-tween the measurements at more than 228 rotational angletoward the film (P , .01) and at each rotational angle,except at 128 rotational angle toward the focal spot (P ,.05). The difference was less than 1% from 08 to 6108rotational angle, and the maximum magnification was1.51% at 1158 rotational angle.

• Posterior facial heightThe posterior facial height was 85.86 mm at 08. The dif-ference between the measurements from 08 to each rota-tional angle was a little larger toward the focal spot thantoward the film, but the degree of these changes was min-imal regardless of the direction of rotation. There was no

statistical difference in the measurements from 08 to eachrotational angle toward the film (P . .05), whereas a sta-tistical difference was found in the measurements at 198,1148 and 1158 rotational angle toward the focal spot (P, .05). The difference was less than 0.5% from 08 to eachrotational angle regardless of the direction of rotation.2. Angular measurements

• SNASNA was 80.868 at 08. There was a statistical difference inthe measurements at 288, 298, 2118 and 2138 rotationalangle toward the film (P , .05), but no statistical differenceexisted in the measurements from 08 to each rotational an-gle toward the focal spot (P . .05). The difference wasless than 0.5% from 08 to each rotational angle regardlessof the direction of rotation.

• SNBSNB was 81.38 at 08. There was no statistical difference inthe measurements from 08 to each rotational angle towardthe film (P . .05), but a statistical difference was found inthe measurements at 1118 and from 1138 to 1158 rota-tional angle toward the focal spot (P , .05). The difference

400 YOON, KIM, HWANG, KIM, CHOI, KIM

Angle Orthodontist, Vol 71, No 5, 2001

TABLE 2. Comparison of the Measurements From Zero to Each Rotational Angle Toward the Focal Spota

Measurement 08 118 128 138 148 158 168

Linear, mm

1. N-S

2. Go-Me

3. N-Me

4. S-Go

MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %

71.274.660.00

77.136.260.00

118.588.810.00

85.869.540.00

71.294.720.03

77.186.330.07

118.69*8.900.10

85.889.520.03

71.234.740.06

77.37*6.200.31

118.718.850.12

86.009.510.17

71.344.750.10

77.61**6.310.62

118.88*8.840.26

85.959.390.11

71.384.800.15

77.64**6.440.66

118.88*8.610.26

85.899.500.04

71.304.790.04

77.46*6.260.43

119.09***8.930.44

86.019.590.18

71.294.720.03

77.61***6.290.63

119.29***9.010.61

86.029.600.19

Angular, 8

5. SNA

6. SNB

7. N-S-Ar

8. S-Ar-Go

9. Ar-Go-Me

10.AB/Go-Me

MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %MeanSDError, %

80.864.980.00

81.306.940.00

133.218.130.00

135.408.130.00

116.596.140.00

75.927.940.00

80.904.910.06

81.396.940.12

133.168.04

20.04135.45

8.130.03

116.656.120.06

75.877.80

20.08

80.825.03

20.0481.446.840.17

133.377.960.12

135.538.010.10

116.606.160.01

75.69*7.75

20.30

80.925.010.09

81.396.800.11

133.267.830.04

135.738.090.25

116.706.190.10

75.707.79

20.29

80.794.94

20.0881.306.970.00

133.298.010.06

135.828.180.31

116.77*6.240.16

75.68**7.80

20.32

80.855.000.00

81.336.990.03

133.288.230.05

135.827.990.31

116.755.960.14

75.63*7.86

20.39

80.804.93

20.0781.326.900.03

133.397.870.14

136.038.040.46

116.846.030.22

75.787.58

20.19

was less than 0.5% from 08 to each rotational angle re-gardless of the direction of rotation.

• Saddle angleThe saddle angle was 133.218 at 08. There was no statisticaldifference in the measurements from 08 to each rotationalangle regardless of the direction of rotation (P . .05). Thedifference was less than 0.5% from 08 to each rotationalangle regardless of the direction of rotation.

• Articular angleThe articular angle was 135.408 at 08. There was a statisticaldifference in the measurements at more than 298 rotationalangle toward the film and more than 188, except at 1118rotational angle toward the focal spot (P , .05). The dif-ference was less than 1.0% from 08 to each rotational angleregardless of the direction of rotation.

• Gonial angleThe gonial angle was 116.598 at 08. There was a statisticaldifference in the measurements at 248, 258 and 298 ro-tational angle toward the film and at 148, 178, and from1118 to 1158 rotational angle toward the focal spot (P ,.05). The difference was less than 1.0% from 08 to eachrotational angle regardless of the direction of rotation.

• AB to mandibular plane angle

AB to mandibular plane angle was 75.928 at 08. There wasno statistical difference in the measurements from 08 toeach rotational angle toward the film (P . .05), but a sta-tistical difference existed in the measurements 128, 148,158, 178 and from 1108 to 1158 rotational angle towardthe focal spot (P , .05). The difference was less than 1.0%from 08 to each rotational angle regardless of the directionof rotation.

DISCUSSION

The resultant images are magnified, because x-rays donot radiate parallel to the whole part of the projected object.The ratio of magnification varies in the different planes,and hence the image is distorted. In the present study, themeasurement values are different according to the directionof rotation because different planes have different magni-fications with different ratios.

Head rotation can occur in the anteroposterior axis, ver-tical axis, and transverse axis. However, standardization onthe anteroposterior and transverse axes is actually hard toachieve. Rotation on the transverse axis causes no distortionof images. Although the head rotates on the transverse axis,

401HEAD ROTATION AND LATERAL CEPHALOMETRIC RADIOGRAPHS

Angle Orthodontist, Vol 71, No 5, 2001

TABLE 2. Extended.

178 188 198 1108 1118 1128 1138 1148 1158

71.154.78

20.1777.43*

6.290.39

119.46***8.940.75

86.179.580.36

71.214.72

20.0877.62**6.240.64

119.57***8.910.84

86.099.440.27

71.02**4.75

20.3577.46*6.320.43

119.72***8.900.97

86.17*9.500.37

70.89***4.73

20.5377.45*6.240.42

119.79***9.011.03

86.029.580.21

70.77***4.68

20.7077.176.380.06

119.95***9.011.17

86.099.510.28

70.59***4.65

20.9577.016.01

20.15119.99***

9.021.20

86.169.560.35

70.42***4.58

21.1976.956.16

20.23120.16***

9.111.34

86.039.540.20

70.11***4.63

21.6376.73*6.15

20.52120.30***

9.061.46

86.19*9.600.39

70.00***4.62

21.7876.38***6.10

20.97120.36***

9.081.51

86.27*9.500.48

80.754.92

20.1481.316.970.02

133.497.870.21

135.858.090.32

117.04**6.120.39

75.64*7.78

20.38

80.745.04

20.1581.216.96

20.11133.42

7.930.16

136.37**8.050.71

116.866.170.03

75.627.69

20.40

80.895.080.05

81.166.93

20.17133.34

7.660.10

136.28*7.880.66

116.896.340.27

75.667.75

20.35

80.894.920.04

81.217.01

20.11133.33

7.830.09

136.44**7.750.78

117.016.200.36

75.53*7.83

20.52

80.835.02

20.0481.13*7.04

20.21133.38

8.080.13

136.158.190.55

117.09**6.390.43

75.52*7.77

20.53

80.845.13

20.0281.107.00

20.25133.28

8.030.05

136.23*7.750.61

117.09*6.150.43

75.49*7.72

20.57

80.785.02

20.1081.07*7.08

20.28133.40

8.050.15

136.51**7.820.81

116.96*6.140.31

75.34*7.62

20.77

80.775.06

20.1181.02*7.02

20.35133.17

7.9620.03136.64**

7.600.92

117.15*6.230.48

75.39*7.76

20.70

80.715.29

20.1880.94*7.05

20.43133.08

7.9620.10136.75**

7.601.00

117.36**6.220.66

75.40*7.65

20.70

a Reference group: 00.* P , .05; ** P , .01; *** P , .001.

the location of the head is parallel to the central ray. Onlythe location of the images on the film changes, but this doesnot cause a change of the relationship between landmarks.1,2

However, rotation in the anteroposterior axis affects land-marks vertically, not horizontally. The bilateral structuresare moved equally and the vertical distance between land-marks changes depending on the distances of landmarksfrom the rotational axis.

Rotation on the vertical axis influenced the horizontalmeasurements, not the vertical measurements, in a differentmanner from rotation on the anteroposterior axis. Unlesslandmarks are located within the distance equidistant fromthe midsagittal plane, any rotation on this axis changes therelationships between the midsagittal line and bilaterallandmarks. Therefore, bilateral landmarks equally placedagainst the midsagittal plane within the skull should bemeasured to remove the adverse effects of rotation on thevertical axis.1,2

Concerning the projection errors of linear measurements,Ahlqvist et al17 supposed that the effects of rotations on theanteroposterior and vertical axes may be identical. In their

study using a computer model similar to the real dry skull,they found that rotation of 658 from the ideal position re-sulted in errors of less than 1%, a margin that is usuallyinsignificant and difficult to distinguish from other errors.However, the errors became significant at even a few de-grees of rotation more than 658. The projection error ofthe present study was different depending on the directionof rotation, contrary to the results of Ahlqvist et al.17

The anterior cranial base length and mandibular bodylength gradually decreased as the rotational angle towardthe film increased. There was a statistical difference in themeasurements from 08 to each rotational angle toward thefilm in the anterior cranial base length (P , .05) and tomore than 228 rotational angle toward the film in the man-dibular body length (P , .01). The difference was above1% with an increase of more than 258 rotational angle, andthe maximum reduction was 25.37% in the anterior cranialbase length and 25.78% in the mandibular body length at2158 rotational angle. These were the same as the resultsof Ahlqvist et al.17 This is thought to result because thenearer to the film the head rotates, the more the image de-

402 YOON, KIM, HWANG, KIM, CHOI, KIM

Angle Orthodontist, Vol 71, No 5, 2001

creases, and this decrease is gradual because the rotationitself causes the decrease of the images.

However, the anterior cranial base length and mandibularbody length increased and then decreased as the rotationalangle toward the focal spot increased. The difference wasless than 1% from 08 to 1128 rotational angle toward thefocal spot in the anterior cranial base length and all rota-tional angles toward the focal spot in the mandibular bodylength, contrary to the results of Ahlqvist et al.17 This isthought to be due to an offset, because the farther from thefilm the head rotates, the more the image is magnified, andthe rotation itself causes the shortening of the images. Thatis, if the image magnification caused by the greater rotationfrom the film is larger than the image reduction caused bythe rotation itself, the length may be increased, but in thereverse case, it will be decreased.

In the anterior facial height, there was a statistical dif-ference in the measurements at more than 228 rotationalangle toward the film (P , .01) and at each rotational an-gle, except at 128 rotational angle toward the focal spot (P, .05). The difference was above 1% at more than 6108rotational angle. In the posterior facial height, the differencewas less than 0.5% at all rotational angles regardless of thedirection of rotation. The variation due to head rotation wasthe least among all the linear measurements.

The amount of rotational axis error was greater in thehorizontal linear measurements than in the vertical linearmeasurements, especially in the mandibular body length.This may be because the landmarks of the mandibular bodylength are located farther vertically from the central ray andcontain bilateral structures. In the case of the posterior fa-cial height including bilateral structures, however, the dif-ference was minimal, because it is near the rotational axis.

Concerning the projection errors of angular measure-ments, Ahlqvist et al18 made various geometric computermodels to find projection errors of angular measurementsand rotated them on the anteroposterior and vertical axes.They demonstrated that rotations within 6108 of the mod-eled angles give rise to angle distortion less than 60.68.Thus, projection errors were insignificant as compared tothe total error. In addition, in their study, computer modelssimilar to the real dry skull were used. When heads wererotated 58 toward both the film and the focus, each distor-tion value was not above 618. Moreover, most of the dis-tortion values were less than 60.58. When 108 rotation wasmade toward the film and the focus, the distortion in-creased, and the value was just within 628.

In summary, for the lateral cephalometric radiograph, 108of head rotation to the film and the focus produced thelargest value possible. In reality, even 58 of head rotationis hard to find in a clinical procedure. The projection erroris insignificant when compared with the total of all the dif-ferences involved in the process. In the present study, SNA,SNB, and saddle angle showed less than 0.5% differenceat all rotational angles regardless of the direction of rota-

tion. Even the more distorted articular angle, gonial angle,and AB to mandibular plane angle showed less than 1%difference, and the projection errors of angular measure-ments were much smaller than those of Ahlqvist et al.18

The projection errors of angular measurements did not ex-ceed 1% difference at all rotational angles regardless of thedirection of angle, and it was far less than those of thelinear measurements.

In conclusion, angular measurements of lateral cephalo-metric radiographs are more useful than linear measure-ments to minimize the projection errors associated withhead rotation on the vertical axis. For example, SUM—acombination of saddle angle, articular angle, and gonial an-gle—has greater diagnostic merit than facial height ratio asone of the linear measurements in the Jarabak Analysis.

Even a tiny error may change the prediction and the anal-ysis of craniofacial growth in a lateral cephalometric radio-graph. Considering that a projection error of 1% differencein the horizontal linear measurements is close to the 1.0mm annual growth19 of the anterior cranial base length(CC-N) and is half of the 2.0 mm annual growth19 of themandibular body length (Xi-pm), these projection errorsshould not be ignored even though they are very small.

Furthermore, considering the projection error caused byhead rotation on the anteroposterior and transverse axes aswell as the vertical axis, the lateral cephalometric radio-graph may lead to an interpretation that is very differentfrom the real condition of the patient. Therefore, in expos-ing films, the projection errors should be reduced as muchas possible. The location of the patient’s head should berepresented consistently. In order to predict and analyze thechange caused by orthodontic treatment and growth, thefilm should be exactly processed and should be accompa-nied by further development of head-positioning devices.

CONCLUSIONS

Angular measurements have fewer projection errors thanlinear measurements. The greater number of landmarks onthe midsagittal plane that are included in angular measure-ments, the fewer the projection errors occurring.

Horizontal linear measurements decrease gradually inlength as the rotational angle toward the film increases,whereas a small increase and then decrease of the lengthoccurs as the rotational angle toward the focal spot increas-es. Horizontal linear measurements have more projectionerrors than vertical linear measurements according to headrotation.

Projection errors of vertical linear measurements increaseas the distance from the rotational axis increases.

In summary, angular measurements of lateral cephalo-metric radiographs are more useful than linear measure-ments in minimizing the projection errors associated withhead rotation on a vertical axis.

403HEAD ROTATION AND LATERAL CEPHALOMETRIC RADIOGRAPHS

Angle Orthodontist, Vol 71, No 5, 2001

ACKNOWLEDGMENTS

The authors would like to thank Drs Pil-Suk Yu, Young-HwanKim, and Seong-Jun Kwon, Residents in Orthodontics, College ofDentistry, Chosun University, for their willing assistance in this in-vestigaion.

REFERENCES

1. Ahlqvist J, Eliasson S, Welander U. The cephalographic projec-tion, II: principles of image distortion in cephalography. Dento-maxillofac Radiol. 1983;12:101–108.

2. Eliasson S, Welander U, Ahlqvist J. The cephalographic projec-tion, I: general consideration. Dentomaxillofac Radiol. 1982;11:117–122.

3. Na KC, Yoon YJ, Kim KW. A study on the errors in the cepha-lometric measurements. Korean J Orthod. 1998;28:75–84.

4. Baumrind S, Frantz RC. The reliability of head film measure-ments, I: landmark identification. Am J Orthod. 1971;60:111–127.

5. Baumrind S, Frantz RC. The reliability of head film measure-ments, II: conventional angular and linear measures. Am J Orthod.1971;60:505–517.

6. Carlsson GE. Error in X-ray cephalometry. Odontol Tidskr. 1967;75:99–129.

7. Cooke MS, Wei SHY. A comparative study of southern Chineseand British Caucasian cephalometric standards. Angle Orthod.1989;2:131–138.

8. Finlay L. Craniometry and cephalometry: a history prior to theadvent of radiography. Angle Orthod. 1980;50:312–321.

9. Graber TM. Implementation of the Roentgenographic cephalo-metric technique. Am J Orthod. 1958;12:906–932.

10. Kantor ML, Phillips CL, Proffit WR. Subtraction radiography toassess reproducibility of patient positioning in cephalometrics.Am J Orthod Dentofacial Orthop. 1993;4:350–354.

11. McWilliam JS, Welander U. The effect of image quality on theidentification of cephalometric landmarks. Angle Orthod. 1978;48:49–56.

12. Midtgard, J Bjork, G Linder-Aronson. S. Reproducibility of ceph-alometric landmarks and errors of measurements of cephalometriccranial distances. Angle Orthod. 1974;44:56–62.

13. Slagsvold O, Pedersen K. Gonial angle distortion in lateral headfilms: a methodologic study. Am J Orthod. 1971;71:554–564.

14. Tng TH, Chan CK, Cooke MS, Orth D, Hagg U. Effect of headposture on cephalometric sagittal angular measures. Am J OrthodDentofacial Orthop. 1993;104:337–341.

15. Buschang PH, Tanguay R, Demirjian A. Cephalometric reliabil-ity: a full ANOVA model for the estimation of true and errorvariance. Angle Orthod. 1987;2:168–175.

16. Houston WJB. The analysis of errors in orthodontic measure-ments. Am J Orthod. 1983;83:382–390.

17. Ahlqvist J, Eliasson S, Welander U. The effect of projection errorson cephalometric length measurements. Eur J Orthod. 1986;8:141–148.

18. Ahlqvist J, Eliasson S, Welander U. The effect of projection errorson angular measurements in cephalometry. Eur J Orthod. 1988;10:353–361.

19. Ricketts RW, Roth RH, Chaconas SJ. Orthodontic Diagnosis andPlanning. Vol. 1. Denver, Colorado: Rocky Mountain Data Sys-tems; 1985.