| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Heilig Hart Hospital, Kolveniersvest 20, 2500 Lier,
Belgium.
2 Department of Pediatrics, Heilig Hart Hospital, 2500 Lier, Belgium.
3 School of Public Health, Biostatistical Centre, Katholieke Universiteit
Leuven, University Hospital St. Rafael, Leuven, Belgium.
Received April 30, 2004;
accepted after revision August 26, 2004.
Address correspondence to T. H. Mulkens
(tom.mulkens{at}hhzhlier.be).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. In standard radiographic examinations of sinuses (anteroposterior and lateral views) with 75 kV, 20 mAs, and 3-mm aluminum filtration in 69 children (mean age, 4.2 years), the dose-area-product (DAP; mGy x cm2) was measured and converted to effective dose (mSv) according to coefficients published by the British National Radiological Protection Board. Another group of 125 children (mean age, 6.8 years) underwent low-dose MDCT of the sinuses with 6- or 16-MDCT in two phases and with different scanning protocols. An effective dose for MDCT was calculated from conversion of the dose-length-product (DLP, mGy xm) according to age.
RESULTS. The mean effective dose (E) for standard radiography was 0.0528 mSv. The mean E value for low-dose MDCT was 0.096 mSv in the first phase of the study but could be lowered in the second phase to 0.0531 mSv by a combination of higher pitch and faster scan rotation time in our scan protocols, which results in diagnostic image quality at a very low dose. Statistical analysis showed no significant differences in effective dose between radiography and MDCT of the second phase.
CONCLUSION. With modern MDCT technology, low-dose CT of the sinuses in children can yield diagnostic image quality using an effective dose comparable to that used for standard radiography.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The increased number of CT examinations also leads to increased X-ray exposure to the public. A recent survey showed that CT now accounts for about 11% of all radiology procedures in the United States and constitutes approximately two-thirds of the collective medical radiation dose [1]. For comparison, CT accounts for only 4% of all radiologic examinations and accounts for more than 40% of the total radiation dose in the United Kingdom [4].
The radiation doses from CT are relatively high and often can approach or exceed the levels known to increase the probability of cancer [3, 5]. Multiple scans present a particular concern: Among children who have undergone CT scans, approximately one-third have had at least three scans [3]. In recent years, there has been an increased concern over the relationship of the use of CT and higher patient and population doses [5–7], especially with regard to the risk for children [3, 7–10]. This issue has highlighted the need for reduction of the radiation dose due to CT.
Sinusitis is a common problem in pediatric practice, since 5–10% of the upper respiratory infections in childhood are complicated by sinusitis [11]. Although radiography is not routinely indicated in these patients, it is still frequently used for diagnosis. Radiography of the sinuses in children is technically demanding, and the interpretation is difficult; there is also a lack of specificity and sensitivity, largely related to the small size of the sinuses, the angulation of the X-ray beam, and nasal secretions [11, 12]. There is a large discrepancy between plain radiographic findings and CT, both in adults and in children, thus, an area of significant controversy in the imaging of sinusitis in children remains the choice between using conventional radiography and using CT [13]. In adults, CT has become the gold standard for sinus imaging, and low-dose CT has thus become the method of choice [14].
We evaluated here the effect of lowering the dose in CT of the sinuses in children to an effective dose that is comparable to the effective dose of a radiographic examination, and we evaluated the image quality obtained at this dose level.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Effective dose conversion factors, that is, the effective dose normalized to DAP (mSv/Gy x cm2), were derived from report NRPB-R 279 of the British National Radiological Protection Board [15], in which coefficients were measured, based on Monte Carlo calculations, in five pediatric phantoms representing neonates and children at 1, 5, 10, and 15 years of age, according to the recommendations of ICRP Publication 60 [16]. Accordingly, we divided our patients into five groups according to age, that is, less than 1 year (n = 10), 1–3 years (n = 23), 4–7 years (n = 23), 8–12 years (n = 11), and greater than 12 years (n = 2), and used the corresponding coefficients of the NRPB-R 279 report for children at birth and at 1, 5, 10, and 15 years of age for each projection to calculate the effective dose.
CT Examinations
From July 1 to December 31, 2003, low-dose CT examination of the sinuses
was performed in 62 children (37 boys and 25 girls; age range, 2–15
years [mean age, 6.5 years]) on a 6- or a 16-MDCT system (Emotion 6 and
Sensation 16, both from Siemens). The following scan protocols were used. For
6-MDCT (n = 30), we used 80 kV, 6-by-0.5-mm collimation, a 1-sec
rotation, a table feed of 4.5 mm per rotation (pitch factor of 1.5), and 20 to
30 effective mAs (according to age). The doses, according to age, were as
follows: less than 3 years, effective mAs of 20, which gives a volume CT dose
index (CTDI vol) of 1.68 mGy; 3–6 years, effective mAs of 25, which
gives a CTDI vol of 2.13 mGy; and greater than 6 years, effective mAs of 30,
which gives a CTDI vol of 2.52 mGy [AQ3].
For 16-MDCT (n = 32), we used 80 kV, 16-by-0.75-mm collimation, a 1-sec rotation, a table feed of 9 mm per rotation (pitch factor of 0.75), and 37 to 45 mAs. The doses, according to age, were as follows: less than 3 years, effective mAs of 37, which gives a CTDI vol of 3.34 mGy; 3–6 years, effective mAs of 40, which gives a CTDI vol of 3.60 mGy; and greater than 6 years, effective mAs of 45, which gives a CTDI vol of 4.00 mGy.
Overlapping images were reconstructed from the raw data set with a 0.63-mm slice thickness and a 0.3-mm increment for the 6-MDCT and a 1-mm slice thickness and a 0.5-mm increment for the 16-MDCT, with a sharp (bone) filter algorithm. From this MDCT data set, consecutive axial and coronal images were obtained with a 2- or 2.5-mm (children > 6 years) slice thickness, with bone window settings (window width, 2,500 H; window center, 500 H), which were used for diagnosis. Sometimes, when needed by the radiologist, sagittal images were reconstructed for better evaluation of the sphenoid sinuses and adenoid volume.
In a second part of our study, from January 1 to April 20, 2004, low-dose CT examination of the sinuses was performed in 63 children (31 boys and 32 girls; age range, 2–16 years [mean age, 7.0 years]) with adapted scan protocols to further lower the radiation dose. The scan rotation was reduced to 0.6 and 0.5 sec per rotation for the 6-MDCT and 16-MDCT, respectively, allowing us to lower the effective mAs. For the 16-MDCT, the pitch factor was put higher to 1.5. According to age, this leads to the following scan protocols. For the 6-MDCT (n = 27), the scan protocols were, according to patient age, as follows: less than 3 years, effective mAs of 15, which gives a CTDI vol of 1.28 mGy; 3–6 years, effective mAs of 20, which gives a CTDI vol of 1.68 mGy; and greater than 6 years, effective mAs of 25, which gives a CTDI vol of 2.12 mGy. For the 16-MDCT (n = 36), the scan protocols were, according to patient age, as follows: less than 3 years, effective mAs of 17, which gives a CTDI vol of 1.43 mGy; 3–6 years, effective mAs of 20, which gives a CTDI vol of 1.68 mGy; and greater than 6 years, effective mAs of 25, which gives a CTDI vol of 2.1 mGy. The same image reconstructions were used as in the first phase. In our study, no child who underwent CT examination of the sinuses underwent radiographic examination of the sinuses and vice versa.
The effective dose (mSv) for the CT scan examinations was calculated by using the method of Chapple et al. [17]. These authors describe a method for correlating the risk-related quantity effective dose with the more simply derived quantity dose-length product (DLP). They found an exponential relationship between the effective dose and the DLP when they scanned a series of pediatric anthropomorphic phantoms containing thermoluminescent dosimeters. We used the equations of Chapple et al. to calculate the effective dose, which is derived from the conversion of DLP, which is given on the scan display of our scanners, at the end of the examination.
The large majority of our children (n = 106; 85%) were referred for CT examination of the sinuses for evaluation of chronic or recurrent complaints: recurrent upper airway infections, chronic nose discharge or nasal congestion, cough (particularly nocturnal), headache, open-mouth breathing, and halitosis. About 15% of our children were referred to CT for evaluation of an acute history with fever, sinus discomfort, or headache or for evaluation of fever of unknown origin.
Informed consent was obtained for all examinations by the pediatrician, who ordered the CT examination and who explained to the parents the use of low-dose scan protocols in our CT examinations.
The study protocol was approved by the ethical committee of our institution. In only two patients were repeat scans necessary; this was because motion artifacts excluded diagnostic image interpretation of the first scan. These patients were excluded from the study.
Statistical Methodology
A Kruskal-Wallis test was used to compare effective doses among the three
different patient groups (radiography, CT phase 1, and CT phase 2
[Fig. 1A]) and among the five
data groups of our study (radiography, 6-MDCT phases 1 and 2, and 16-MDCT
phases 1 and 2 [Fig. 1B]), and
Wilcoxon's tests were used for pairwise comparisons of data groups. A
significance level alpha of 5% was used. A Bonferroni correction was applied
for the paired equations.
|
|
Evaluation of Image Quality
Radiological images are interpreted by human observers; therefore, image
quality in medicine can be judged in terms of the extent to which a class of
images allows observers (e.g., radiologists) to determine correctly whether
each examination is normal or pathologic
[18]. We compared the
performance of two radiologists and one pediatrician in their classification
of the CT examinations as normal or pathologic in order to obtain an
impression of the consistency in diagnosis of the CT examinations of the
sinuses with our low-dose protocols and in this way obtain an impression of
the image quality. One hundred CT examinations were randomly chosen (50 from
CT phase 1 and 50 from CT phase 2) and reviewed on a workstation (Leonardo,
Siemens) by the three viewers independent of each other: two radiologists,
each with more than 10 years of experience in CT, and one pediatrician who was
not experienced in the evaluation of CT examinations. For each sinus a score
was made, normal (score 0) or pathologic (score 1). Pathologic findings
included the following: complete or nearly complete opacification of the
sinus, the presence of an air-fluid level and/or mucosal thickening of at
least 4 mm. The presence of adenoid hypertrophy was scored positive if the
nasal airway was completely obstructed or narrowed to a distance of 3 mm or
less. The presence of fluid in the middle ear cavities was evaluated and
scored as otitis media (serosa) for each side. In this way, a score range from
zero to a maximum of 11 was possible. Each sinus was scored separately, and
these scores and the total score of the viewers were statistically compared by
using paired Wilcoxon's test between each pair of viewers. A kappa test was
used to evaluate interobserver agreement.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
In phase 1 of the CT study, the mean effective dose for all CT examinations was 0.096 mSv (Table 1): 0.0697 mSv for the 6-MDCT and 0.1218 mSv for the 16-MDCT. The mean DLP for both examinations was 24.57 mGy x cm: 17.7 mGy x cm for the 6-MDCT and 31 mGy x cm for the 16-MDCT. The mean scanning times of the examination were 16.9 sec for the 6-MDCT and 6.9 sec for the 16-MDCT. The mean scan lengths were 7.68 cm for the 6-MDCT and 7.96 cm for the 16-MDCT.
In phase 2 of the CT study, the mean effective dose for all CT examinations was lowered to 0.0531 mSv (Table 1): 0.0511 mSv for the 6-MDCT and 0.0546 mSv for the 16-MDCT. The mean DLP for both CT examinations was 13.9 mGy x cm: 12.9 mGy x cm for the 6-MDCT and 14.7 mGy x cm for the 16-MDCT. The mean scanning time was lowered to 10.1 sec for the 6-MDCT and to only 2.1 sec for the 16-MDCT. The mean scan length was 7.25 cm for the 6-MDCT and 7.90 cm for the 16-MDCT.
Statistical analysis with a Kruskal-Wallis test showed a significant difference in effective dose anong the three groups together (p < 0.0001).
Pairwise comparison with Wilcoxon's tests gave a significant difference in effective dose between the CT phase 1 and radiography groups (p < 0.0001), between the CT phase 1 and CT phase 2 groups (p < 0.0001), and between the CT phase 2 and radiography groups (p = 0.048, near threshold level). The p values alone are not meaningful for evaluating the equivalence or noninferiority of different methods: significant p values do not take into account the clinical relevance of the observed difference, and nonsignificant p values can be due to the lack of power of the study. Therefore, the observed difference and its confidence interval should also be evaluated. Intuitively, the smaller (the closer to zero) the limit of the confidence interval, the more evidence for equivalence. The difference in mean effective dose between the radiography and CT phase 2 groups is 0.0010, with 0.0104 as the upper limit of the 95% one-sided confidence interval, a difference that is negligible.
Further pairwise analysis based on Wilcoxon's tests showed a significant difference in effective dose between the radiography group and each CT phase 1 group (p < 0.01 and p < 0.001) and between 6-MDCT phase 1 and 16-MDCT phase 2 (p < 0.001) but not between the radiography group and each CT phase 2 group (p = 0.22 and 0.06) and between the two MDCT techniques in phase 2 (p = 0.5).
Evaluation of image interpretation with paired Wilcoxon's rank tests and a Kruskal-Wallis test showed no statistical differences between the scoring of the presence of sinusitis signs in 100 CT examinations between each pair of reviewers (p = 0.39, 0.16, and 0.06 for the Wilcoxon's tests) and among the three reviewers together (p = 0.88 for the Kruskal-Wallis test). The Spearman's correlation coefficients between the scores of the pediatrician and the two radiologists were 0.94 and 0.95, respectively, and 0.97 between the two radiologists. Kappa coefficients for interobserver agreement were 0.96 between both radiologists and 0.87 between each radiologist and the pediatrician. Refined analysis of the scores showed that all 17 negative CT examinations were scored negative by the three reviewers (Figs. 2A, 2B and 3A, 3B). There were identical scores of 85/100 for the two radiologists and 67/100 for the three reviewers together. In another three examinations there was an identical score among the three observers, but this was due to positive scoring of different sinuses with the same total score as result. Differences in score were especially due to a discrepancy in the evaluation of the sphenoid sinuses in young children and in the evaluation of the middle ears. In children 6 years old or younger, the sphenoid sinuses are small (Fig. 4A, 4B, 4C) or not yet developed. The presence of otitis (Figs. 5A, 5B, 5C and 6A, 6B, 6C) was scored in 15 cases by one radiologist, in 13 cases by the second radiologist, and only five times by the pediatrician. Most of these children had extensive sinusitis signs or bilateral pan sinusitis (Figs. 5A, 5B, 5C and 6A, 6B, 6C).
|
|
|
|
|
|
|
|
|
|
|
|
|
Compared with the "default" examination protocols for CT examination of the sinuses in children, as proposed by the manufacturer, the dose used in our low-dose protocols, expressed in CTDI vol, is five to seven times lower.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT of the sinuses involves imaging of tissue with three widely different densities: air, bone, and soft tissue. This gives an inherent high image contrast, which allows the possibility of using low-dose CT with good diagnostic image quality [13, 21]. Low-dose CT of the sinuses has already been used for many years in adults [22, 23] and, with the introduction of MDCT technology, which obviates the use of direct coronal scanning, the use of low-dose CT has become the method of choice because of comparable diagnostic image quality in comparison with standard-dose CT [21]. Even at the lowest mAs settings of 16 and 23 mAs in the axial and coronal planes at 120 kV, the degree of soft-tissue visualization necessary to diagnose sinusitis was achieved [23]. The same findings have been confirmed by Tack et al. [21] using 4-MDCT with 120 kV, only 10 mAs, a 4-by-1-mm collimation, and a pitch factor of 2. The calculated effective dose in both of these studies for adults is in the same range as the calculated effective dose for the children in our study, that is, 0.05 to 0.12 mSv.
The use of conventional radiography versus (coronal) CT for imaging of sinusitis in the pediatric population continues to be controversial [13]. Standard radiography lacks sensitivity and specificity, and a radiography-based evaluation can either over- or underestimate soft-tissue changes in the paranasal cavities [9, 12]. The use of plain films for the diagnosis of sinusitis should be discouraged [13]. Although the low sensitivity and specificity of standard radiographic examination of the sinuses is well known and proven in comparative studies with CT in both adults and children [12, 22], the use of CT in the evaluation of the sinuses in children has remained limited because of the higher radiation dose of CT (especially for the eye lens and thyroid), the need for sedation of preschool children, and the higher cost [11]. Because of low cost and wide availability, radiographic examinations are likely to continue to play an important role in the diagnosis and follow-up of children with medically managed sinusitis [11]. In the head and neck region, the eye lens and thyroid gland are the most radiosensitive organs [16, 24]. The threshold dose for detectable lens opacities is 0.5–2 Gy for acute exposure or 150 mGy per year [16, 24]. In a study of sinus CT, Zammit-Maempel et al. [24] used 4-MDCT with 140 kV, 100 mAs, and 1-mm collimation, with resulting mean doses of 35.1 mGy for the lens and 2.9 mGy for the thyroid gland. In a low-dose protocol with 120 kV and 20 mAs, these authors measured doses of 9.2 mGy and 0.4 mGy for lens and thyroid, which is much less than the threshold doses [24]. With the children in the present study, we used low-dose CT protocols at even lower dose settings.
We found no reports in the current literature on the use of low-dose CT of the sinuses in a pediatric population. After implementation of a new 6-MDCT system in our department in February 2003, we were not completely satisfied with the image quality of the "default" scan protocol offered by the manufacturer. We altered the settings to a low-dose level. The image quality was very good at this lower dose, and so we changed the protocol for CT of the sinus in children on our 16-MDCT system in the same way. With our latest protocols (phase 2), we pushed our CT machines to their lowest possible dose levels (minimum of 80 kV, lowest possible mAs settings, and high pitch factor) but still achieved sufficient image quality to make an accurate diagnosis. With the use of wide window settings and reconstruction of thicker sections (2- or 2.5-mm adjacent images) from submillimeter collimation MDCT scans, the inherent noise of low-dose CT was largely reduced in our scan protocols and was at an acceptable level.
In cases of extensive sinus disease, the thickness and integrity of the bone septa (e.g., ethmoid septa and lamina papyracea) are sometimes difficult to evaluate on low-dose CT images [14]. This was also the case in our study. Thus, low-dose CT can be used to confirm the clinical diagnosis of sinusitis and to evaluate its extent. Low-dose CT of the sinuses in children is a good alternative to the use of conventional radiography because of its higher accuracy at a comparable effective dose. Low-dose CT techniques lack soft-tissue contrast: when sinusitis complications or possible underlying malignant disease must be evaluated, a higher CT dose, the use of an IV contrast agent, or alternative imaging with MRI should be used [14].
A drawback of our study is that we did not compare the image quality of our
low-dose CT protocols with the gold-standard, normal-dose CT of the sinuses.
In a low-dose CT study of sinusitis in adults, Hagtvedt et al.
[25] reached a very high
specificity (
96%) and sensitivity ( 95%, except for the frontal sinus
[83%]). Since children have less-dense calcified bones, the CT image quality
would be expected to be better at a comparable (low) dose. We used a scoring
system to assess image quality and obtained excellent interobserver agreement,
reflecting the good sensitivity of the low-dose technique.
A second controversy in imaging pediatric sinusitis is the high incidence of soft-tissue changes in the sinus cavities found in radiographic, CT, and MR images in children who have no clinical evidence of sinus disease and who undergo medical imaging for other clinical reasons. This incidence is reported to be 33–50% [26–29]. A common cold (or other upper airway infection) acutely produces mucosal abnormalities in sinuses in the majority of adults and children [28], especially in patients who had a cold in the 2 weeks preceding imaging. Therefore, the diagnosis of acute and chronic or recurrent sinusitis should not be made based on the imaging findings alone [13, 28]. A diagnosis of acute or chronic sinusitis should be made clinically, with confirmation based on laboratory and imaging findings [13].
In conclusion, our study shows that low-dose CT of the sinuses offers a good alternative to standard radiographic examination. Low-dose CT using a comparable effective dose shows good image quality, follows the ALARA concept in pediatric CT, and can replace plain radiography in the evaluation of children clinically suspected of having acute or chronic sinus disease.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  U.K. Udayasankar, K. Braithwaite, M. Arvaniti, D. Tudorascu, W.C. Small, S. Little, and S. Palasis Low-Dose Nonenhanced Head CT Protocol for Follow-Up Evaluation of Children with Ventriculoperitoneal Shunt: Reduction of Radiation and Effect on Image Quality AJNR Am. J. Neuroradiol., April 1, 2008; 29(4): 802 - 806. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Just da Costa e Silva and G. Alves Pontes da Silva Eliminating Unenhanced CT When Evaluating Abdominal Neoplasms in Children Am. J. Roentgenol., November 1, 2007; 189(5): 1211 - 1214. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
