Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Geijer, H.
Right arrow Articles by Persliden, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Geijer, H.
Right arrow Articles by Persliden, J.
(Radiology. 2001;218:402-410.)
© RSNA, 2001

Medical Physics

Digital Radiography of Scoliosis with a Scanning Method: Initial Evaluation1

Håkan Geijer, MD, BA, Karl-Wilhelm Beckman, MSc, Berith Jonsson, RN, Torbjörn Andersson, MD, PhD and Jan Persliden, PhD

1 From the Departments of Radiology (H.G., T.A.) and Medical Physics (K.W.B., B.J., J.P.), Örebro Medical Centre Hospital, 701 85 Örebro, Sweden. From the 1999 RSNA scientific assembly. Received February 8, 2000; revision requested April 7; revision received May 25; accepted July 31. Supported by the research committee of Örebro County Council and the Swedish Radiation Protection Institute, project SSI P1141.99. Address correspondence to H.G. (e-mail:hakan.geijer@orebroll.se).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
PURPOSE: To evaluate the radiation dose, image quality, and Cobb angle measurements obtained with a digital scanning method of scoliosis radiography.

MATERIALS AND METHODS: Multiple images were reconstructed into one image at a workstation. A low-dose alternative was to use digital pulsed fluoroscopy. Dose measurements were performed with thermoluminescent dosimeters in an Alderson phantom. At the same time, kerma area-product values were recorded. A Monte Carlo dose calculation also was performed. Image quality was evaluated with a contrast-detail phantom and visual grading system. Angle measurements were evaluated with an angle phantom and measurements obtained on patient images.

RESULTS: The effective radiation dose was 0.087 mSv for screen-film imaging, 0.16 mSv for digital exposure imaging, and 0.017 mSv for digital fluoroscopy; the corresponding kerma area-product values were 0.43, 0.87, and 0.097 Gy · cm2, respectively. The image quality of the digital exposure and screen-film images was about equal at visual grading, whereas fluoroscopy had lower image quality. The angle phantom had lower angle values with digital fluoroscopy, although the difference in measured angles was less than 0.5°. The patient images showed no difference in angles.

CONCLUSION: The described digital scanning method has acceptable image quality and adequate accuracy in angle measurements. The radiation dose required for digital exposure imaging is higher than that required for screen-film imaging, but that required for digital fluoroscopy is much lower.

Index terms: Dosimetry • Images, quality, 30.11, 30.1215 • Radiations, exposure to patients and personnel • Radiography, digital, 30.1215 • Spine, curvature, 30.861 • Spine, radiography, 30.11, 30.1215


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Scoliosis radiography has generally been performed by using a screen-film system with a large cassette. In a digital environment with use of storage phosphor screens, scoliosis radiography is difficult to perform because there are no detector plates large enough to cover the entire spine, with the exception of the spine of small children. Scoliosis radiography has been performed also by using image intensifiers but with the same type of size restrictions. Previously described techniques (1) to overcome this problem usually have been performed with dedicated equipment.

In diagnostic radiology, a low radiation dose based on the ALARA (as low as reasonably achievable) principle (2) is important. In scoliosis radiography, this principle is especially important, because patients with scoliosis are young and often subjected to repeated studies owing to long treatment periods. Most evaluations of scoliosis radiography in the literature have included vague estimates of the radiation doses used.

The digital method described by van Eeuwijk et al (3) helps overcome the problem of detector size. Briefly, this examination consists of a multiple-image scan of the entire spine. The images are then reconstructed into one image at a workstation. A variant of this method is grid-controlled pulsed fluoroscopy.

The aim of our study was to investigate the radiation dose, image quality, and accuracy of Cobb angle measurements obtained with the digital scanning method compared with those obtained with the conventional screen-film method.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
For the parts of the study that involved additional imaging in patients, informed consent was obtained. In addition, the local research ethics committee approved this part of the study. All other patient imaging studies were performed as routine examinations.

This study was designed to compare three methods of scoliosis radiography—two digital techniques and one screen-film method. The two digital methods—digital exposure imaging and digital fluoroscopy—are similar in many ways and were performed with the equipment described in the next section. The screen-film method was evaluated as a reference standard for comparison with the digital methods.

Radiographic Techniques
All the digital imaging (ie, digital exposure and digital fluoroscopy) examinations were performed with a Multi Diagnost 4 (Philips Medical Systems, Eindhoven, the Netherlands) digital image intensifier–based unit with capabilities for digital exposure and grid-controlled pulsed fluoroscopy. Equipment data are shown in Table 1. The equipment was used in the factory setting, and all methods were evaluated the way they were set during installation.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Imaging Parameters and Equipment
 
For the digital examinations, the patient stands erect and faces the image intensifier in a posteroanterior projection (Fig 1). The entire spine is covered by using a series of images synchronized with a craniocaudal scan of the x-ray tube and image intensifier C arm. The radiation field for each image should, according to the manufacturer, be collimated to about 10 cm in the craniocaudal direction at the input screen of the image intensifier. In the lateral direction, the collimation is adjusted to the degree of scoliosis to yield a field size for each image of 20–30 x 10 cm at the input screen of the image intensifier. At our institution, the imaged field covers the area from the occiput to the hips, a total of about 70 cm in the craniocaudal direction.



View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Alderson dose measurement phantom faces the image intensifier in a posteroanterior projection.

 
For digital exposure imaging, the tube potential (kilovolt peak level) is fixed and the tube current (milliamperage) is varied automatically during scanning. With a frequency of two images per second and a scanning speed of nominally 4 cm/sec, 30–40 images are obtained. The other digital method, digital fluoroscopy, involves grid-controlled pulsed fluoroscopy during C-arm movement in which a series of images are saved to memory. The tube potential and tube current are adjusted automatically during scanning. Because of a higher pulse frequency, three images per second, 45–60 images are obtained with the same scanning speed as that used for digital exposure imaging.

Lateral views are not routinely used at our institution and thus were not included in this study. However, these views can be obtained in the same way posteroanterior projections are obtained.

All images are exported to a workstation with commercially available software (EasyVision; Philips Medical Systems). The images are merged and reconstructed into one overview image in less than 30 seconds with the aid of pattern recognition (Fig 2). The borders of the individual images cannot be seen on overview images. An overview image has a size of, for example, 512 x 1,300–2,000 pixels. With the same software, measurements of Cobb angles, femur height difference, and vertical alignment can be performed easily. Finally, annotated images can be printed out or, as at our institution, exported to a picture archiving and communicating system (PACS) (Sectra and Philips Medical Systems, Linköping, Sweden). All images are viewed on a monitor, and image manipulations, such as gray-scale correction, magnification, and measurement, can be performed at any workstation in the hospital that is connected to the PACS.



View larger version (142K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Reconstructed frontal digital exposure image (posteroanterior projection) of the spine that consists of data from 30-40 images that were reconstructed into one image. The image on the left shows the individual images that were merged to form the image on the right.

 
For comparison purposes, a conventional screen-film examination, performed in a posteroanterior projection in an erect patient, was evaluated. The radiated field was about 25 x 70 cm, and no compensating filters were used. In 1997, our radiology department moved to a new building with completely new digital radiography equipment; thus, the screen-film method is no longer used at our institution. Because the equipment routinely used for imaging with film had been discarded at the time of this study, the dose measurements were performed with other similar equipment (Table 1). All settings were chosen to be as similar as possible to those of the old equipment. The settings for the new equipment were verified by imaging the dose-measurement phantom with film to yield a correctly exposed image. The images acquired with the three methods are shown in Figure 3.



View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. A, Frontal screen-film image of the spine of a 15-year-old girl, which was used as the reference standard for image quality assessment. B, On the frontal digital exposure image obtained in an 18-year-old woman, the metal artifact overlying the lower lumbar spine is a piece of jewelry in the pierced navel. C, Frontal digital fluoroscopic image obtained in the same patient and on the same day as in B.

 
Dose Measurements
Phantom measurements obtained with thermoluminescent dosimeters (TLDs).—The effective radiation dose was calculated from absorbed organ doses measured in an Alderson-type humanoid phantom (4,5) that represented a female person with a height of 163 cm and weight of 54 kg and was loaded with TLDs (TLD-100; Harshaw Chemical, Solon, Ohio). The TLD disks were placed in the relevant organs to calculate the effective radiation dose (6). Several TLDs were placed dorsally to record the entrance dose values as a basis for the Monte Carlo calculations. The TLDs were individually calibrated and read with a TLD reader (Alnor Dosacus; Alnor Oy, Turku, Finland).

The TLD-loaded phantom was irradiated at three separate sessions for each method; reading and reloading were performed between each session to assess absorbed dose variability. With screen-film and digital exposure imaging, the phantom was examined five times at each irradiation event to increase the radiation exposure levels enough to obtain reliable TLD readings. With digital fluoroscopy, we chose to examine the phantom 20 times at each irradiation event for the same reason. With digital fluoroscopy, a higher pulse frequency (three pulses per second) was used because this was the lowest frequency available.

Kerma area-product values.—Kerma area-product values were measured by using a transmission ion chamber connected to an electrometer (Doseguard 100; RTI Electronics, Mölndal, Sweden).

Monte Carlo calculations.—With the Monte Carlo method, a computer simulation of the experimental situation was performed. For all three methods, the organ doses and effective dose were calculated with the PCXMC computer program (Finnish Centre for Radiation and Nuclear Safety, Helsinki, Finland) (7). For the two digital methods, a separate calculation had to be performed for each image; this resulted in 37 calculations for digital exposure imaging and 58 calculations for digital fluoroscopy. The results of these separate calculations were added to obtain the results for each method. All Monte Carlo calculations were tailored to be applicable for a 15-year-old patient of the same size as the phantom.

Image Quality
Contrast-detail phantom.—The CDRAD 2.0 contrast-detail phantom (Instrumentale Dienst, Nijmegen, the Netherlands) (8) was used to perform a semiobjective evaluation of image quality. This phantom is a 1-cm-thick slab of acrylic with drilled holes of different sizes and depths. The number of holes visible in each column can be drawn as a graph, which gives a depiction of image quality. A numeric value, the image quality figure, also can be calculated. A low image quality figure means better image quality. The phantom was placed between two 7.5-cm-thick acrylic sheets to simulate the scattering conditions of the human body. The phantom was imaged once with each method. Each reconstructed image was read once by three observers (including H.G. and T.A.) independently.

Visual grading.—The image quality of the patient images was evaluated by using visual grading (9) in which each image was compared with a reference screen-film image. An image database was formed from 30 screen-film examinations, performed in 30 patients (six male, 24 female; mean age, 13.8 years; age range, 7–19 years), from 1995 that were sequentially retrieved from the archives; 30 digital exposure images, obtained in 30 patients (six male, 24 female; mean age, 13.9 years; age range, 8–18 years), from 1998 to 1999; and 10 digital fluoroscopy images, obtained in 10 female patients (mean age, 15.0 years; age range, 11–18 years), from 1998 to 1999. Ten of these patients, after giving informed consent, underwent a double examination with both digital exposure imaging and digital fluoroscopy. Because it was impossible to perform imaging in the same patient with all three methods, the study was performed in two steps.

Thirty screen-film and 30 digital exposure images were compared with one reference image, which was chosen as a representative screen-film scoliosis radiograph of good quality. Each image was evaluated separately and independently by the three observers.

The following levels were evaluated: third thoracic vertebra (T3), 10th thoracic vertebra (T10), and third lumbar vertebra (L3). These levels were chosen to focus the observers on smaller parts of the image. On the old screen-film images, the uppermost level was often overexposed, whereas the two lower levels usually were correctly exposed.

The following aspects of image quality were evaluated: contrast level, noise (ie, quantum mottle), sharpness, visualization of vertebral end plates, subjective suitability for judging skeletal abnormalities, and subjective suitability for angle measurements. Furthermore, the overall quality of the image was assessed.

At each level and aspect, image quality was independently scored on a five-level scale as much worse, worse, equal, better, or much better compared with the reference image quality. The images obtained in the 10 patients who underwent double examinations (a total of 20 images) were graded in pairs, with the digital exposure image in the pair used as the reference. The same grading aspects as those described earlier were used.

This scoring system resulted in nearly 4,000 gradings. All digital images were evaluated at a PACS workstation by using a 1,280 x 1,024-pixel gray-scale monitor (Image Systems, Hopkins, Minn).

Angle Measurements
Angle phantom.—The accuracies of Cobb angle measurements obtained with the three methods were evaluated with a phantom made of 13.5 cm of acrylic with 2-mm-thick aluminum sheets simulating two vertebrae in the middle (Fig 4). The aluminum sheets could be rotated between images. Thirty different angles between 0° and 53° were randomly set and imaged with all three methods. Each angle was independently measured once by all three observers. The digital images were measured with the standard PACS software; the film images were marked with a pencil and measured with a protractor. All observers used the same protractor during the entire study. All lines were erased before leaving the image.



View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Photograph (A) and radiographic image (B) of the phantom used for the angle measurements. The phantom is composed of acrylic, with aluminum sheets simulating vertebrae. The diagonal wire mesh was necessary for proper reconstruction of the image.

 
Clinical images.—The images used for visual grading were also used for angle measurements. Each angle was measured twice by the three observers. All measurements were made independently, and the double measurements were made at different sessions. Which levels to use for measurement was decided beforehand because this affects measurement accuracy (10). Images were measured the same way as described earlier.

Statistical Methods
The visual grading material was presented descriptively. Agreement between observers was evaluated with {kappa} statistic analysis; verbal interpretations of the {kappa} values are shown in Table 2. The angle measurements were evaluated with two-way analysis of variance.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Verbal Interpretation of {kappa} Values
 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Dose Measurements and Image Quality
When the scanning speed of the equipment was checked, it ran at a speed of about 3.5 cm/sec instead of the nominal 4 cm/sec. This naturally increased the overlap between images and thus the radiation dose.

The results of measurements and calculations in the Alderson phantom are shown in Table 3. Variability between irradiation events was low. The large difference in dose between digital exposure imaging and fluoroscopy should be noted. For comparison, the median kerma area-product value for 20 lumbar spine examinations (four to five images each at 70–90 kVp) at our institution is 6.6 Gy · cm2.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Measured and Calculated Phantom Dose Values
 
The digital exposure variant yielded an image quality figure of 64, which is considerably better than 88, the image quality figure for digital fluoroscopy, but worse than 47, the image quality figure for screen-film imaging. The contrast-detail curves are shown in Figure 5.



View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Graph illustrates measurement curves for the contrast-detail phantom. The mean values obtained by the three observers are depicted. The further down to the left the curve lies, the better the image quality. {blacktriangleup} = digital fluoroscopy; image quality value = 88. {blacklozenge} = digital exposure imaging; image quality value = 64. {blacksquare} = screen-film imaging; image quality value = 47.

 
Visual Grading
Screen-film versus reference image.— The reference image was overexposed in the area of the upper part of the thoracic spine. This was reflected in the judging of all the screen-film images as better than the reference image in all aspects except noise at the T3 level. At the two lower levels, T10 and L3, all screen-film grades were somewhat lower than those on the reference image, except those for noise, which was graded as equal. The overall quality of the screen-film images was slightly worse than that of the reference image.

{kappa} statistic analysis results showed moderate to substantial agreement between observers ({kappa}, 0.46–0.76) on the clinical aspects of visualization of vertebral end plates, suitability for judging skeletal abnormalities, and suitability for angle measurements. For image contrast, noise, and sharpness, agreement was lower, with {kappa} values down to -0.03 (poor agreement). However, with regard to the aspects with the lowest {kappa} values, the observers agreed in 87 of 90 gradings; this showed the limitation of {kappa} statistic analysis.

Digital exposure versus reference image.—Compared with the overexposed reference image, at the T3 level, the digital exposure images were judged to be better in all aspects except noise, which was graded as somewhat lower. The two lower levels, T10 and L3, were similar in all aspects, with lower grades on the digital exposure images. The overall quality of the digital exposure images was better.

With regard to visualization of vertebral end plates, suitability for judging skeletal abnormalities, and suitability for angle measurements, there was less agreement between observers at digital exposure image grading; agreement ranged from slight to substantial ({kappa}, 0.14–0.65).

Digital exposure versus screen-film images.—Compared with the grades at the T3 level on the screen-film images, those on the digital exposure images were better in all aspects except noise. At the T10 and L3 levels, which had similar grades, contrast, noise, and sharpness were graded somewhat lower on the digital exposure images, but visualization of end plates, suitability for judging skeletal abnormalities, and suitability for angle measurements were graded somewhat higher. The overall quality of the digital exposure images was considerably better. Gradings at the T10 level on screen-film and digital exposure images are illustrated in Figure 6. The results at the L3 level were similar.



View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6a. Graphs illustrate the visual grading of the screen-film and digital exposure images compared with the reference screen-film image—performed on 30 images for a total of 90 gradings of each aspect—by three observers. The gradings of (a) contrast at the T10 level, (b) suitability for judging skeletal abnormalities at the T10 level, (c) suitability for angle measurements at the T10 level, and (d) overall quality are depicted.

 


View larger version (31K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6b. Graphs illustrate the visual grading of the screen-film and digital exposure images compared with the reference screen-film image—performed on 30 images for a total of 90 gradings of each aspect—by three observers. The gradings of (a) contrast at the T10 level, (b) suitability for judging skeletal abnormalities at the T10 level, (c) suitability for angle measurements at the T10 level, and (d) overall quality are depicted.

 


View larger version (32K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6c. Graphs illustrate the visual grading of the screen-film and digital exposure images compared with the reference screen-film image—performed on 30 images for a total of 90 gradings of each aspect—by three observers. The gradings of (a) contrast at the T10 level, (b) suitability for judging skeletal abnormalities at the T10 level, (c) suitability for angle measurements at the T10 level, and (d) overall quality are depicted.

 


View larger version (34K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6d. Graphs illustrate the visual grading of the screen-film and digital exposure images compared with the reference screen-film image—performed on 30 images for a total of 90 gradings of each aspect—by three observers. The gradings of (a) contrast at the T10 level, (b) suitability for judging skeletal abnormalities at the T10 level, (c) suitability for angle measurements at the T10 level, and (d) overall quality are depicted.

 
Digital fluoroscopy versus digital exposure images.—At all levels and in all aspects, digital fluoroscopy was graded lower than digital exposure imaging in the same patient, with the least difference being in suitability for angle measurements. The overall quality grade of digital fluoroscopic images also was lower than that of digital exposure images. The results are shown in Table 4.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Visual Grading of Digital Fluoroscopy versus Digital Exposure Imaging at the T10 Level
 
The observers had moderate agreement on contrast and sharpness at all levels ({kappa} = 0.46), and there was complete agreement on noise among the observers, who all chose the same grade—worse. An intelligible {kappa} value could not be calculated with this total agreement. With regard to suitability for judging skeletal abnormalities, agreement was moderate to almost perfect ({kappa}, 0.46–1.00). Agreement on visualization of vertebral end plates and suitability for angle measurements was poor.

Angle Measurements
Angle phantom.—Two-way analysis of variance revealed significant results for the two factors of method and observer (method: F = 5.8, P = .005; observer: F = 6.4, P = .003; method and observer: F = 2.3, P = .062). For the post-hoc test, a 95% CI of plus or minus 0.38° was calculated. By using this CI, there were significant differences between observers 1 and 3 and between observers 2 and 3 for digital fluoroscopy. Likewise, there were significant differences between screen-film imaging and fluoroscopy and between digital exposure imaging and fluoroscopy for observer 3. Neither of the other comparisons was significant. Observer 3 and digital fluoroscopy yielded smaller angle measurements than did the other two techniques. Mean angle measurements are shown in Table 5 and Figure 7; the differences between angle measurements obtained with the different methods and the mean angle measurement are shown in Figure 8.


View this table:
[in this window]
[in a new window]

  		TABLE 5. Means and SDs of Cobb Angle Measurements Obtained in a Phantom
 


View larger version (33K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7. Graph illustrates mean phantom angle measurements obtained by the different observers with each method. Slightly smaller angles were measured with digital fluoroscopy. {blacksquare} = observer 1, {blacklozenge} = observer 2, {blacktriangleup} = observer 3.

 


View larger version (27K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8. Graph illustrates differences in phantom angle measurements from the mean angle measurement obtained with the different methods. For each angle, the results obtained by all the observers were averaged. The small measurement error occurred because the phantom images permitted far more exact measurements than do patient images. The depicted measurements show that measurement error is not dependent on angle size. The fluoroscopy angle measurements were slightly lower than the mean angle measurements. {blacksquare} = difference in screen-film versus mean angle measurement, {circ} = difference in digital exposure versus mean angle measurement, + = difference in digital fluoroscopy versus mean angle measurement.

 
Clinical images.—When comparing the ten patients who underwent double examinations with digital exposure imaging and fluoroscopy, there was no significant difference in angle measurements (method: F = 1.1, P = .32; observer: F = 0.61, P = .55; method and observer: F = 0.45, P = .64). The mean angles and reproducibility of angle measurements are shown in Table 6. All images with an angle difference larger than 5° between the highest and lowest measured value (two screen-film, four digital exposure, and no digital fluoroscopy images) were scrutinized for possible causes of the large variation. On most images, simply poor definition of the vertebral end plates caused a large part of guessing in placing the measuring lines. Two such patients—one who underwent screen-film imaging; and one, digital exposure imaging—were previously operated on with metal rods along the spine. One patient had several vertebral malformations. One patient, who underwent digital exposure imaging, had 13 thoracic vertebrae; because of this, one of the reviewers measured the angle at the wrong level. This angle was subsequently measured again at the correct level. Examples of images with large measurement errors are shown in Figure 9.


View this table:
[in this window]
[in a new window]

  		TABLE 6. Means and SDs of Cobb Angle Measurements Obtained with the Different Imaging Techniques
 


View larger version (46K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9a. Images with a large error in the measurement of the Cobb angle. (a) Frontal radiograph with 5.4° between the highest and lowest measurements in a 16-year-old boy who previously underwent surgery. (b) Frontal radiograph with 7.4° between the highest and lowest measurements in a 14-year-old boy with a large rotation in the lumbar spine.

 


View larger version (46K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9b. Images with a large error in the measurement of the Cobb angle. (a) Frontal radiograph with 5.4° between the highest and lowest measurements in a 16-year-old boy who previously underwent surgery. (b) Frontal radiograph with 7.4° between the highest and lowest measurements in a 14-year-old boy with a large rotation in the lumbar spine.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
REFERENCES
 
The digital exposure method for scoliosis radiography makes radiography of long objects possible in a digital environment, even when formats of storage phosphor screens and image intensifiers are too small to accommodate imaging with a single exposure. The effective dose of 0.16 mSv used with digital exposure imaging is higher than the 0.087 mSv used with our previously used screen-film method. When pulsed digital fluoroscopy is used, however, the effective dose, 0.017 mSv, is considerably lower.

The advantages of digital exposure imaging are that all images have a consistent exposure along the entire length of the spine without under- or overexposed areas. Angle measurements are easy to perform with dedicated software at the workstation. This method also works for lateral scoliosis radiographs, but great care has to be taken to get acceptable exposure levels. The resultant images can be stored both with and without annotation on a PACS or printed. Use of the large and unwieldy full-size film used in the conventional method is thus avoided.

Some problems are that the examination is somewhat complicated to perform, and experienced technicians are required for good results. Collimation is a constant problem, because any increase in field size in the craniocaudal direction results in a proportional increase in dose. A fixed collimator might be a solution. Images acquired with the digital exposure method have a large overlap, and each point of the patient is exposed by at least two images. According to the manufacturer, this overlap is necessary for accurate reconstruction of the overview image. The fact that the scanning speed is slower than nominal points to the importance of acceptance testing when installing new equipment. A setup outside the stated specifications might result in a higher-than-anticipated radiation dose.

The imaging geometry with this digital method is unusual, because in the craniocaudal direction the x-ray beam is largely parallel. In the lateral direction, however, it has the traditional fan-beam shape. This causes, at least in theory, the potential for image distortion when objects are at different distances from the receptor (1). This theory was not investigated further in our study.

Radiation Dose
The radiation dose in scoliosis radiography has previously been lowered in various ways (11–13). Changes in the screen-film system at conventional imaging can reduce the dose considerably (14). Posteroanterior instead of anteroposterior projection reduces the dose to the breast (15) and other sensitive organs at the cost of a higher bone marrow dose (16). Posteroanterior projection has been reported to be superior in image quality to anteroposterior projection (17), which is now considered obsolete in scoliosis radiography (18). An undercouch tube position, when used with a tabletop, is valuable because the tabletop attenuates considerably. We have measured attenuation as high as 30% with the parameters used in this study. Such an attenuator between the patient and the image receptor should be avoided because this increases the dose to the patient. Increasing x-ray beam filtration is another way to lower the dose (13). With the transition to digital imaging, a further decrease in radiation dose is expected (19,20).

To our knowledge, calculations of the effective dose in scoliosis radiography are scarce in the literature. Values of 0.12 mSv (21) (frontal and lateral views) and 0.05 mSv (22) (frontal view) for conventional screen-film systems have been reported. The dose measurement phantom used in this study was developed for measurements in radiation therapy with higher photon energies (4). The attenuation properties in the diagnostic energy range for this phantom differ somewhat from those for human tissue (23). For the heavily filtered x-ray beam we use, this difference is small, with water having a linear attenuation coefficient that is only 1% lower. The ease of use and general availability of this type of phantom made us choose it despite these potential shortcomings.

Monte Carlo calculations have been used previously in the scoliosis radiography setting, but without any values of effective dose given (15). Our Monte Carlo calculations showed a remarkably good correlation with the measured values of effective dose, even in this complicated setting with multiple images. This indicates that Monte Carlo calculations can be a useful tool in the clinical setting for evaluating the effective dose in various examination types.

Kerma area-product values are easy to obtain and can be a useful tool to compare radiation doses among different institutions and different methods. Kerma area-product values are more sensitive to increases in kilovolt peak level and filtration than effective doses and thus might exaggerate the dose gain.

Image Quality
The digital exposure images were of good quality and had better grades than did the screen-film scoliosis images, despite visible noise, especially when they were magnified. This was also reflected in the visual grading part of the study, in which the digital exposure images were rated as superior to the screen-film images in the subjective aspects of visualization of end plates, suitability for judging skeletal abnormalities, suitability for angle measurements, and overall quality, despite having inferior contrast, noise, and sharpness. With the contrast-detail phantom, digital exposure image quality was lower than screen-film image quality; we interpreted this to be mainly the result of a different imaging chain (image intensifier) and the scanning process with parallax errors. It seems that despite having lower objective image quality, digital exposure images are preferred to screen-film images—probably at least in part because of the uneven exposure of the screen-film images (24). The ability to alter the gray scale and contrast dynamically at a PACS workstation also is a great advantage. Digital fluoroscopy images had lower quality and marked noise and were clearly inferior to digital exposure images both at contrast-detail phantom readings and visual grading.

Angle Measurements
There was no significant difference in angle phantom measurements between screen-film and digital exposure images. There was, however, a significant difference between digital fluoroscopy and the other two methods. The mean angles were 21.6° at both screen-film and digital exposure imaging and 21.2° at digital fluoroscopy, with a much lower mean angle SD on the angle phantom, about 0.5°, compared with about 1.0° on the patient images. This indicated a considerably higher precision in the artificial measuring situation, which means that even a small difference between methods turned out to be significant. The difference in mean angles between the three observers was somewhat larger with digital fluoroscopy. It must be stressed, however, that no significant difference between measurements was found on the patient images. The reproducibility, expressed in intraobserver SD, was good with all three methods and better than that reported elsewhere (25). We believe the small difference in angle phantom measurements would not be clinically important in real-life practice, because differences of up to 10° between measurements have been required to indicate a real change in the scoliosis angle (26). The interobserver, intraobserver, and patient differences were much larger. We consider both digital methods to be acceptable for measurements of scoliosis angles, with the caveat that digital fluoroscopy might be slightly less exact. Further studies in this area are needed.

One way to exploit the advantages of the higher image quality of digital exposure imaging and the lower radiation dose with digital fluoroscopy might be to combine the two methods (19). Digital exposure imaging could be used for the initial examination, in which it is important to evaluate underlying abnormalities, whereas digital fluoroscopy could be used for the follow-up examinations, in which the focus is angle measurements.

Statistical Methods
Some authors have set a numeric value to each level (ie, much worse, worse, equal, etc) in the visual grading scale and calculated mean values (9). We disagree with this methodology because the levels are ordinal data with no assumption of equal steps between the different levels and thus should not be treated as continuous data. Instead, we chose to give a descriptive presentation of our material. The maximum {kappa} value is 1.00 when agreement is perfect, whereas a value of zero indicates no agreement better than chance. {kappa} statistic analysis is of limited value when most of the gradings for all observers accumulate at only one level (eg, are equal). Then, calculations yield low {kappa} values despite excellent agreement. In the extreme case when all the grades on all the images for all observers are identical, the {kappa} value cannot be calculated owing to a division by zero.

In conclusion, in this study, we evaluated a method for digital scoliosis radiography that enables one to overcome the size restriction of storage phosphor screens. Digital exposure image quality is acceptable, but the radiation dose should be lowered further. A low-dose alternative is to use pulsed digital fluoroscopy, which involves a substantial dose reduction with acceptable accuracy in angle measurements. We are planning future studies in which the described digital exposure examination is optimized further, particularly with lower radiation doses.


   ACKNOWLEDGMENTS
 
The authors thank Frida Högberg, BA, and Leif Norén, BA, of the Department of Statistics of Örebro University for statistical assistance, radiographer Lena Evaldsson, RN, for excellent imaging, and Robert Popek, MD, for many hours of image quality assessment.


   FOOTNOTES
 
Abbreviations: PACS = picture archiving and communicating system, TLD = thermoluminescent dosimeter

Author contributions: Guarantor of integrity of entire study, H.G.; study concepts, H.G., B.J.; study design, H.G., T.A., J.P.; definition of intellectual content, H.G., T.A., J.P.; literature research, H.G.; clinical studies, H.G.; experimental studies, T.A., H.G., B.J., K.W.B.; data acquisition, H.G., B.J.; data analysis, H.G., K.W.B.; statistical analysis, H.G., J.P.; manuscript preparation, H.G.; manuscript editing, H.G.; manuscript review and final approval, all authors.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 

  1. Kushner DC, Cleveland RH, Herman TE, Zaleske DJ, Ehrlich MG, Correia JA. Radiation dose reduction in the evaluation of scoliosis: an application of digital radiography. Radiology 1986; 161:175-181.[Abstract/Free Full Text]
  2. International Commission on Radiological Protection. 1977 recommendation of the International Commission on Radiological Protection. ICRP Publication 26 Oxford, United Kingdom: Pergamon, 1977.
  3. van Eeuwijk AHW, Lobregt S, Gerritsen FA. A novel method for digital x-ray imaging of the complete spine. In: Troccaz J, eds. Lecture notes in computer science: proceedings of CVRMed-MRCAS ’97, March 19–22, 1997. Grenoble, France. Berlin, Germany: Springer-Verlag, 1997; 1205:521-530.
  4. Alderson SW, Lanzl LH, Rollins M, Spira J. An instrumented phantom system for analog computation of treatment plans. AJR Am J Roentgenol 1962; 87:185-195.
  5. International Commission on Radiation Units and Measurements. Phantoms and computational models in therapy, diagnosis and protection. ICRU Report no. 48 Bethesda, Md: International Commission on Radiation Units and Measurements, 1992; 48.
  6. International Commission on Radiological Protection. 1990 recommendation of the International Commission on Radiological Protection. ICRP Publication no. 60 Oxford, United Kingdom: Pergamon, 1991; 21.
  7. Tapiovaara M, Lakkisto M, Servomaa A. PCXMC: a PC-based Monte Carlo program for calculating patient doses in medical x-ray examinations: version 1.3, STUK-A139 Helsinki, Finland: Finnish Centre for Radiation and Nuclear Safety, 1997.
  8. Thijssen MAO, Thijssen HOM, Merx JL, Lindeijer JM, Bijkerk KR. A definition of image quality: the image quality figure. In: Moores BM, Wall BF, Eriskat H, Schibilla H, eds. BIR Report 20: optimization of image quality and patient exposure in diagnostic radiology. London, United Kingdom: British Institute of Radiology, 1989; 29-34.
  9. Manninen H, Terho EO, Wiljasalo M, Wiljasalo S, Soimakallio S. An evaluation of different imaging chains in clinical chest radiography. Br J Radiol 1984; 57:991-995.[Abstract]
  10. Morrissy RT, Goldsmith GS, Hall EC, Kehl D, Cowie GH. Measurement of the Cobb angle on radiographs of patients who have scoliosis: evaluation of intrinsic error. J Bone Joint Surg Am 1990; 72-A:320-327.[Abstract/Free Full Text]
  11. Nash Jr CL, Gregg EC, Brown RH, Pillai K. Risks of exposure to x-rays in patients undergoing long-term treatment for scoliosis. J Bone Joint Surg Am 1979; 61:371-374.[Abstract/Free Full Text]
  12. Ardran GM, Coates R, Dickson RA, Dixon-Brown A, Harding FM. Assessment of scoliosis in children: low dose radiographic technique. Br J Radiol 1980; 53:146-147.[Medline]
  13. Gray JE, Hoffman AD, Peterson HA. Reduction of radiation exposure during radiography for scoliosis. J Bone Joint Surg Am 1983; 65:5-12.[Abstract/Free Full Text]
  14. Hellström G, Irstam L, Nachemson A. Reduction of radiation dose in radiologic examination of patients with scoliosis. Spine 1983; 8:28-30.[Medline]
  15. Levy AR, Goldberg MS, Mayo NE, Hanley JA, Poitras B. Reducing the lifetime risk of cancer from spinal radiographs among people with adolescent idiopathic scoliosis. Spine 1996; 21:1540-1547.[Medline]
  16. Fearon T, Vucich J, Butler P, et al. Scoliosis examinations: organ dose and image quality with rare-earth screen-film systems. AJR Am J Roentgenol 1988; 150:359-362.[Abstract/Free Full Text]
  17. Frank ED, Stears JG, Gray JE, Winkler NT, Hoffman AD. Use of the posteroanterior projection: a method of reducing x-ray exposure to specific radiosensitive organs. Radiol Technol 1983; 54:343-347.
  18. Almén A, Lööf M, Mattsson S. Examination technique, image quality, and patient dose in paediatric radiology: a survey including 19 Swedish hospitals. Acta Radiol 1996; 37:337-342.[Medline]
  19. Kalmar JA, Jones JP, Merritt CRB. Low-dose radiography of scoliosis in children: a comparison of methods. Spine 1994; 19:818-823.[Medline]
  20. Stringer DA, Cairns RA, Poskitt KJ, Bray H, Milner R, Kennedy B. Comparison of stimulable phosphor technology and conventional screen-film technology in pediatric scoliosis. Pediatr Radiol 1994; 24:1-5.[Medline]
  21. Hallén S, Martling K, Mattsson S. Dosimetry at x-ray examinations of scoliosis. Radiat Prot Dosimetry 1992; 43:49-54.[Abstract]
  22. Almén AJ, Mattsson S. Dose distribution at radiographic examination of the spine in pediatric radiology. Spine 1996; 21:750-756.[Medline]
  23. Shrimpton PC, Wall BF, Fisher ES. The tissue-equivalence of the Alderson Rando anthropomorphic phantom for x-rays of diagnostic qualities. Phys Med Biol 1981; 26:133-139.[Medline]
  24. Jónsson Á, Jonsson K, Eklund K, Holje G, Pettersson H. Computed radiography in scoliosis: diagnostic information and radiation dose. Acta Radiol 1995; 36:429-433.[Medline]
  25. Manninen H, Kiekara O, Soimakallio S, Vainio J. Reduction of radiation dose and imaging costs in scoliosis radiography: application of large-screen image intensifier photofluorography. Spine 1988; 13:409-412.[Medline]
  26. Carman DL, Browne RH, Birch JG. Measurement of scoliosis and kyphosis radiographs: intraobserver and interobserver variation. J Bone Joint Surg Am 1990; 72:328-333.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 RadioGraphicsHome page
M. Korner, C. H. Weber, S. Wirth, K.-J. Pfeifer, M. F. Reiser, and M. Treitl
Advances in Digital Radiography: Physical Principles and System Overview
RadioGraphics, May 1, 2007; 27(3): 675 - 686.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Radiol.Home page
M Bath and L G Mansson
Visual grading characteristics (VGC) analysis: a non-parametric rank-invariant statistical method for image quality evaluation
Br. J. Radiol., March 1, 2007; 80(951): 169 - 176.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Geijer, H.
Right arrow Articles by Persliden, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Geijer, H.
Right arrow Articles by Persliden, J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE