RESEARCH—HUMAN—CLINICAL STUDIES RESEARCH—HUMAN—CLINICAL STUDIES
Strategies for Computed Tomography Radiation Dose Reduction in Pediatric Neuroimaging Gregory W. Albert, MD, MPH* Charles M. Glasier, MD‡ *Division of Neurosurgery, ‡Department of Radiology, Arkansas Children’s Hospital and University of Arkansas for Medical Sciences, Little Rock, Arkansas Correspondence: Gregory W. Albert, MD, MPH, Division of Neurosurgery, Arkansas Children’s Hospital, 1 Children’s Way, Slot 838, Little Rock, AR 72202. E-mail: [email protected] Received, November 19, 2014. Accepted, February 24, 2015. Published Online, April 7, 2015. Copyright © 2015 by the Congress of Neurological Surgeons.
BACKGROUND: Radiation exposure from diagnostic imaging is a significant concern, particularly in the care of pediatric patients. Computed tomography (CT) scanning is a significant source of radiation. OBJECTIVE: To demonstrate that diagnostic quality CT images can be obtained while minimizing the effective radiation dose to the patient. METHODS: In this retrospective cross-sectional study, noncontrast head CT scan data were reviewed, and indications for scans and estimated radiation dose delivered were recorded. The estimated effective radiation dose (EERD) for each CT protocol was reviewed. RESULTS: We identified 251 head CT scans in a single month. Of these, 96 scans were using a low-dose shunt protocol with a mean EERD of 0.82 mSv. The remaining 155 scans were performed using the standard protocol, and the mean EERD was 1.65 mSv. Overall, the EERD was minimized while maintaining diagnostic scan quality. CONCLUSION: Although replacing a CT with magnetic resonance imaging is ideal to completely avoid ionizing radiation, this is not always practical or preferred. Therefore, it is important to have CT protocols in place that minimize radiation dose without sacrificing diagnostic quality. The protocols in place at our institution could be replicated at other academic and community hospitals and imaging centers. KEY WORDS: Child, CT protocol, Neuroimaging, Pediatrics, Radiation dosage, Tomography, X-ray computed Neurosurgery 77:228–232, 2015
DOI: 10.1227/NEU.0000000000000764
R
adiation exposure from medical imaging, especially computed tomography (CT) scanning, is of increasing concern. This is particularly true in the pediatric age group. Cancer risk due to CT scanning was highlighted in a recent publication by Pearce et al1 that reported an increased risk of brain cancer and leukemia in patients with higher lifetime exposure to CT radiation. Multiple methods have been proposed to limit the radiation exposure of children from diagnostic imaging. These techniques include implementation of protocols to limit the number of scans ordered and the replacement of CT scan with other modalities such as magnetic resonance imaging (MRI) and head ultrasound. There are situations, however, when CT is indicated, and in these situations, it is prudent to minimize the radiation dose applied. ABBREVIATION: EERD, estimated effective radiation dose
228 | VOLUME 77 | NUMBER 2 | AUGUST 2015
www.neurosurgery-online.com
Over many years, the neuroradiologists at our institution have developed protocols to minimize the estimated effective radiation dose (EERD) delivered with head CT scanning while maintaining diagnostic utility. The CT protocol selected for a particular patient depends on the indication for the scan and the age of the patient (Table 1). Here we present our age- and diagnosis-specific noncontrast head CT protocols. We evaluate the EERD administered to the patients stratified by indication for the scan and the age of the patient. Similar low-dose protocols could be applied anywhere CT scanning is performed, whether at a large academic referral hospital or at a smaller imaging center in the community.
METHODS For this retrospective cross-sectional study, the radiology records at our hospital were reviewed. We recorded data from all noncontrast head CT scans performed in a single month (October 2013). Volumetric scans performed for surgical planning
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
CT DOSE REDUCTION
TABLE 1. Noncontrast Head Computed Tomography Protocols Used at Our Institution Peak Killivoltage
Tube Current, mA
Time, s
80 100
50-100 50-100
0.75-1 0.75-1
100 100 120 120
80-200 80-230 80-230 80-300
0.75-1 0.75-1 0.75-1 0.75-1
Shunt ,1 y old .1 y old Standard ,2 y old 2-10 y old 10-18 y old .18 y old
(neuronavigation scans) were excluded. The dose length product was recorded during the acquisition of each study. The dose length product was then converted to the estimated effective radiation dose (EERD) in millisieverts using previously published age-based conversion factors (age 0-1 year = 0.011; age 1-5 years = 0.0067; age 5-10 years = 0.0040; age 10-15 years = 0.0032; age 15 and older years = 0.0021).2 Multiple head CT protocols are used at our institution depending on the indication for the scan and the age of the patient (Table 1). Patients for whom the indication was hydrocephalus, ventriculomegaly, or the presence of a cerebrospinal fluid (CSF) shunt were assigned to the “shunt” protocol group. All other patients were assigned to the “standard” protocol group. Because the purpose of this study was to evaluate how the scanning protocols are applied and the radiation received by patients, group assignments were based on the clinical indications for the scan rather than on the protocol used for the image acquisition. The majority of scans also used iterative reconstruction in postprocessing. All patients undergoing cranial imaging only have lead aprons applied to protect their thyroids and bodies. No special protection is applied to the corneas of the patients. The mean and standard deviation of the EERD was calculated for the shunt group patients and the standard group patients. Patients for whom the EERD was .2 standard deviations above the mean were considered outliers, and the individual scans reviewed to determine the reason for the increased radiation exposure. Patients were further stratified by age according to the head CT protocols at our institution, 2 subgroups for shunt protocol patients (younger than 1 year of age, 1 year of age and older) and 4 subgroups for standard protocol patients (younger than 2 years of age, 2-10 years of age, 10-18 years of age, 18 years of age and older). Again, the mean and standard deviation of the EERD were calculated for each subgroup. We randomly selected one-third of the scans for review by a pediatric neurosurgeon and a pediatric neuroradiologist. Each of these scans was scored on a 3-point scale (1 = ideal scan, 2 = not ideal but still diagnostic, 3 = nondiagnostic). All statistical calculations were performed with Microsoft Excel (Microsoft, Redmond, Washington) and StatPlus:mac (AnalystSoft, Inc, Alexandria, Virginia). This study was approved by our institution’s institutional review board.
RESULTS We identified a total of 251 noncontrast head CT scans performed in a single month. Patient ages ranged from 0 to 21
NEUROSURGERY
years. Of these, 96 scans were performed for hydrocephalus/ ventriculomegaly/CSF shunt and were designated as the shunt protocol. The remaining 155 scans were performed for a variety of indications including trauma, headache, macrocephaly, and seizure and were designated as the standard protocol (Figure). For the shunt protocol CT scans (Table 2), the mean EERD was 0.82 mSv with a standard deviation of 0.50 mSv (range, 0.24-3.02 mSv). Eighty-two of the scans (86%) were ,1.0 mSv. Six scans were .2 standard deviations above the mean. In 4 of these instances, the patient was scanned with a tube current that was greater than that specified for the shunt protocol. The other 2 patients were scanned according to protocol. Fifteen shunt protocol scans were performed in patients younger than 1 year of age. The mean EERD for this group was 1.01 mSv with a standard deviation of 0.45 mSv. The remaining 81 scans were in patients older than 1 year of age, and the mean EERD was 0.78 mSv with a standard deviation of 0.50 mSv. The standard protocol CT scans (Table 3) had a mean EERD of 1.65 mSv and a standard deviation of 0.97 mSv (range, 0.546.95 mSv). A total of 134 scans (87%) were ,2.5 mSv. Two scans delivered an EERD of .2 standard deviations above the mean. In both cases, the patient was scanned twice due to extreme motion artifact on the initial scan. Forty-nine standard protocol CT scans were performed in patients younger than 2 years of age. The mean EERD for this group was 2.52 mSv with a standard deviation of 1.25 mSv. Fortyfive patients were 2 to 10 years of age. The mean EERD for these patients was 1.31 mSv with a standard deviation of 0.42 mSv. Fifty-seven patients were 10 to 18 years of age and were exposed to a mean EERD of 1.24 mSv with a standard deviation of 0.38 mSv. The remaining 4 patients were 18 years of age or older. This group received a mean EERD of 0.97 mSv with a standard deviation of 0.16 mSv. We reviewed 84 of the scans for quality. Twenty-five of these scans were from the shunt protocol group and 59 of these scans were from the standard protocol group. This represents one-third of all the scans, 26% of the shunt protocol scans, and 38% of the standard protocol scans. No scan was given a score of 3 (nondiagnostic). Of the shunt protocol scans reviewed, 3 (12%) were scored as 2 (not ideal but still diagnostic). Two of these scans had motion artifact, and 1 did not use iterative reconstruction in postprocessing. Of the standard protocol scans, 2 (3%) were scored as 2. One of these was due to motion artifact and the other due to lack of iterative reconstruction in postprocessing. The Fisher exact test showed no statistically significant difference in the quality scores of the shunt protocol scans compared with the standard protocol scans (P = .15).
DISCUSSION The radiation dose used in CT scans has been of increasing concern in the past several years. There are now population-based clinical data demonstrating an increased cancer risk with increased lifetime exposure to diagnostic radiation,1 although a recently
VOLUME 77 | NUMBER 2 | AUGUST 2015 | 229
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
ALBERT AND GLASIER
FIGURE. Representative images obtained with the reduced radiation CT protocols. A, shunt, age 0.4 years, 0.65 mSv. B, shunt, age 4.8 years, 0.95 mSv. C, standard, age 0.4 years, 2.61 mSv. D, standard, age 5.0 years, 1.38 mSv. E, standard, age 12.1 years, 1.40 mSv. F, standard, age 18.9 years, 0.97 mSv.
published smaller clinical study suggests that this risk is low.3 Many groups have proposed methods of reducing radiation dose. These include replacing CT with MRI and performing low-dose CT scans with either a limited number of slices or reduced radiation dose per slice.
TABLE 2. Estimated Effective Radiation Dose Delivered for Scans Performed With Shunt Protocol (Hydrocephalus/Ventriculomegaly/ Cerebrospinal Fluid Shunt)a EERD, mSv
All studies Age ,1 y Age .1 y a
No.
Mean
SD
96 15 81
0.82 1.01 0.78
0.5 0.45 0.5
EERD, estimated effective radiation dose; SD, standard deviation.
230 | VOLUME 77 | NUMBER 2 | AUGUST 2015
Brain MRI Many groups have described replacing CT with MRI, particularly in nonemergent situations. This approach has been most widely described in patients with hydrocephalus and CSF shunts. Various sequences have been used including SSFSE, T1-FLASH, T2-HASTE, and diffusion-weighted imaging.4-7 We use rapid sequence MRI in our institution for routine hydrocephalus follow-up in cooperative patients. We routinely obtain images in all 3 anatomic planes with a scan time of 1 minute. We generally reserve this approach for patients whom we know can stay still without anesthesia. However, the primary limiting factor at our institution is the availability of MRI. During the daytime, the scanners are frequently occupied, and at night and on weekends, there is no in-house MRI technician. Although urgent scans are possible, there is a time delay that is not practical for emergency situations such as shunt malfunction and trauma. In addition, MRI lacks detail for bone imaging and visualization of shunt catheters. In our practice, it is
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
CT DOSE REDUCTION
TABLE 3. Estimated Effective Radiation Dose Delivered for Scans Performed With Standard Protocola EERD, mSv
All studies Age, y ,2 y 2-10 10-18 .18 a
n
Mean
SD
155
1.65
0.97
49 45 57 4
2.52 1.31 1.24 0.97
1.25 0.42 0.38 0.16
EERD, estimated effective radiation dose; SD, standard deviation.
still preferable or even necessary to use head CT scans for many patients. Limited-Slice CT Scan Two groups recently published studies exploring the use of limited-slice head CT scans as a means of reducing radiation dose in children with CSF shunts. Both studies used existing images from standard or reduced dose head CT scans and created theoretical limited-slice CT scans consisting of either 38 or 79 slices. Although appealing as a means to reduce radiation exposure, experience shows that some patients will have a subtle change in ventricle size that may be missed on limited-slice head CT. In addition, the cases of false-positive findings on limitedslice head CT could have led to unnecessary surgery. Also, limited-slice head CT scanning may not adequately visualize the ventricular catheter or overall ventricular anatomy, both of which are important for surgical planning. Finally, this is an approach that is only applicable to shunt follow-up and not suitable for trauma, altered mental status, macrocephaly, etc. Low-Dose Head CT Scan The most commonly reported approach for limited radiation exposure from medical imaging in children is the use of reduced-dose head CT scanning protocols. Various methods have been reported for decreasing the dose. Most commonly, this is accomplished by reducing the tube current (milliamperes) used for each study.10-12 Adjustments in tube rotation time have also been reported.10 In some cases, postprocessing is also used to reduce the noise in the image and therefore improve image quality.12 Reduced-dose CT scan protocols have also been of interest to craniofacial surgeons.13 Low-dose CT scanning is particularly applicable in these cases because lower radiation doses are needed for high-quality bone imaging than for brain imaging. In all these cases, significant reductions in effective radiation dose were achieved while maintaining the diagnostic utility of the imaging studies. Multiple CT protocols have been developed at our institution that optimize the radiation dose without sacrificing the diagnostic utility of the study (Table 1). Selection of the CT protocol
NEUROSURGERY
depends on both the indication for the scan and the age of the patient. Smaller patients are more susceptible to the effects of radiation, and reduced dose protocols in these children are important. The EERD for our standard protocol is lower than that for other widely used protocols. The shunt protocol radiation dose is even lower, 50% of the standard protocol. The dose reduction is most significant in the younger age groups. Regardless of the dose-reduction protocols, younger children still receive a larger EERD. This is expected given that younger children are more susceptible to the effects of radiation as indicated by the agedependent conversion factors. Overall, the reduced radiation protocols that we use have been successful in minimizing the effective radiation dose delivered to our patients while still allowing us to obtain diagnostic-quality images. As expected, a greater percentage of the shunt protocol scans were of lower quality. Generally, lower image quality is acceptable when the primary concern is evaluating ventricle size vs looking for potentially subtle changes in brain parenchyma. Therefore, there is a lower threshold for repeating the imaging of a standard protocol scan than of a shunt protocol scan. This repetition at the time of scanning is also why no scan in this study was scored as a 3 (nondiagnostic). Those patients with repeated scans were, however, exposed to greater levels of radiation, as noted earlier. We have also found that the use of iterative reconstruction is critical for the maintenance of diagnostic quality with these imaging protocols, as evidenced by the fact that several of the lower quality scans can be attributed to the lack of iterative reconstruction. CT scanners are more widely available than MRI scanners. Although it may be practical for large medical centers to use MRI instead of CT in most situations, this approach is not generalizable to all hospitals. Until such a time that all hospitals have ready access to MRI scanners around the clock, attention must be paid to limiting the radiation dose from CT scanning. The techniques that we use for radiation reduction could be easily applied at other institutions including those with limited access to MRI. Limitations This study is limited by its retrospective nature. In addition, specific information regarding the CT protocol to which individual patients were assigned was not reviewed. This allows for a realworld evaluation of our CT scanning protocols, as the radiation dose delivered to a particular patient may be higher than planned due to misassignment of the protocol, inappropriate scanner settings, or the need to repeat the scan for any reason. Finally, our dosimetry data indicate that our patients are being scanned at EERD lower than the pediatric reference standard levels. Therefore, we had no control group.
CONCLUSION The risk of ionizing radiation to children, particularly the increased risk of malignancy, is a valid concern. Head CT scanning
VOLUME 77 | NUMBER 2 | AUGUST 2015 | 231
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
ALBERT AND GLASIER
is a significant source of radiation for children. Multiple strategies exist for limiting the exposure of children to ionizing radiation from CT scans, including reducing the number of scans performed and replacing CT with MRI in selected instances. There are times that CT is necessary, and in these instances, it is important to minimize the radiation dose without sacrificing diagnostic yield. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Pearce MS, Salotti JA, Little MP, et al. Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet. 2012;380(9840):499-505. 2. Thomas KE, Wang B. Age-specific effective doses for pediatric MSCT examinations at a large children’s hospital using DLP conversion coefficients: a simple estimation method. Pediatr Radiol. 2008;38(6):645-656. 3. White IK, Shaikh KA, Moore RJ, et al. Risk of radiation-induced malignancies from CT scanning in children who underwent shunt treatment before 6 years of age: a retrospective cohort study with a minimum 10-year follow-up. J Neurosurg Pediatr. 2014;13(5):514-519. 4. O’Neill BR, Pruthi S, Bains H, et al. Rapid sequence magnetic resonance imaging in the assessment of children with hydrocephalus. World Neurosurg. 2013;80(6): e307-e312. 5. Wait SD, Lingo R, Boop FA, Einhaus SL. Eight-second MRI scan for evaluation of shunted hydrocephalus. Childs Nerv Syst. 2012;28(8):1237-1241. 6. Ashley WW Jr, McKinstry RC, Leonard JR, Smyth MD, Lee BC, Park TS. Use of rapid-sequence magnetic resonance imaging for evaluation of hydrocephalus in children. J Neurosurg. 2005;103(2 suppl):124-130. 7. Iskandar BJ, Sansone JM, Medow J, Rowley HA. The use of quick-brain magnetic resonance imaging in the evaluation of shunt-treated hydrocephalus. J Neurosurg. 2004;101(2 suppl):147-151. 8. Alhilali LM, Dohatcu AC, Fakhran S. Evaluation of a limited three-slice head CT protocol for monitoring patients with ventriculoperitoneal shunts. AJNR Am J Roentgenol. 2013;201(2):400-405. 9. Pindrik J, Huisman TA, Mahesh M, Tekes A, Ahn ES. Analysis of limitedsequence head computed tomography for children with shunted hydrocephalus: potential to reduce diagnostic radiation exposure. J Neurosurg Pediatr. 2013;12(5): 491-500. 10. Rybka K, Staniszewska AM, Biega nski T. Low-dose protocol for head CT in monitoring hydrocephalus in children. Med Sci Monit. 2007;13(suppl 1):147-151. 11. Udayasankar UK, Braithwaite K, Arvaniti M, et al. 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. 2008; 29(4):802-806. 12. Morton RP, Reynolds RM, Ramakrishna R, et al. Low-dose head computed tomography in children: a single institutional experience in pediatric radiation risk reduction: clinical article. J Neurosurg Pediatr. 2013;12(4):406-410. 13. Vazquez JL, Pombar MA, Pumar JM, del Campo VM. Optimised low-dose multidetector CT protocol for children with cranial deformity. Eur Radiol. 2013; 23(8):2279-2287.
Acknowledgments The authors thank Donna Hoover, BSRT(R)(MR)(CT), for helping to identify the patients and scans for this study and Daniel Harton, BSRT(R)(CT), for providing the CT protocols.
232 | VOLUME 77 | NUMBER 2 | AUGUST 2015
COMMENTS
A
lthough this was a retrospective, single-center study, the authors accumulated a large number of patients over a short time period and managed to demonstrate their premise that diagnostic quality scans can be achieved with reduced radiation dose when CT must be used. I was particularly pleased to see images taken under the lower dose protocol and the standard, demonstrating the preserved quality of the images and the diagnostic effectiveness. Neurosurgeons, especially pediatric neurosurgeons, will want to have this information to assist with patient care. Attention must be given to the details of the technique used by this center to achieve dose reduction with useful scan quality. As they discuss, other strategies used for a similar goal have limitations. Ann Marie Flannery Lafayette, Louisiana
T
he authors document a significant reduction in the estimated effective radiation dose (EERD) for children undergoing CT imaging of the brain using a CSF shunt protocol. The authors only specify protocol-based limitations in peak kilovoltage and tube current, based on age. Despite the noted reductions in EERD, no shunt protocol scan was nondiagnostic, although there was a trend toward more frequent occurrence of “not ideal but still diagnostic” scans in this group (12% for shunt protocol vs 3% for standard protocol scans). In real-life clinical practice, using an “exaggerated sniff” patient position in the CT scanner can achieve substantial additional reduction in EERD, while largely avoiding exposure of the thyroid and corneas to radiation.1 The ideal strategy for radiation exposure reduction during pediatric axial brain imaging is to substitute MRI for CT imaging.2 As the authors point out, however, MRI for emergency indications is not logistically or financially feasible in many centers. In addition, simple strategies are also important: judicious ordering of axial imaging appropriate to the clinical situation, use of radiography rather than CT when clinically appropriate (for example, in most cases of cervical spine clearance after minor trauma), and restricting the use of repeat imaging in trauma for nonevolving pathologies.3 The authors have contributed valuable information to an ongoing effort to reduce the exposure of pediatric patients to CT radiation. This is particularly important for young patients, such as those with hydrocephalus, often exposed to frequent axial brain imaging. Nathan R. Selden Portland, Oregon
1. Didier RA, Kuang AA, Schwartz DL, Selden NR, Stevens DM, Bardo ME. Decreasing the effective radiation dose in pediatric craniofacial CT by changing head position. Pediatr Radiol. 2010;40(12):1910-1917. 2. Yue EL, Meckler GD, Fleischman RJ, et al. Test characteristics of quick brain MRI for shunt evaluation in children: an alternative modality to avoid radiation. J Neurosurg Pediatr. Epub 2015 Jan 30. doi: 10.3171/2014.9.PEDS14207. 3. Durham SR, Liu KC, Selden NR. Utility of serial computed tomography imaging in pediatric head trauma. J Neurosurg. 2006;105(5 suppl):365-369.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
Strategies for Computed Tomography Radiation Dose Reduction in Pediatric Neuroimaging Gregory W. Albert, MD, MPH* Charles M. Glasier, MD‡ *Division of Neurosurgery, ‡Department of Radiology, Arkansas Children’s Hospital and University of Arkansas for Medical Sciences, Little Rock, Arkansas Correspondence: Gregory W. Albert, MD, MPH, Division of Neurosurgery, Arkansas Children’s Hospital, 1 Children’s Way, Slot 838, Little Rock, AR 72202. E-mail: [email protected] Received, November 19, 2014. Accepted, February 24, 2015. Published Online, April 7, 2015. Copyright © 2015 by the Congress of Neurological Surgeons.
BACKGROUND: Radiation exposure from diagnostic imaging is a significant concern, particularly in the care of pediatric patients. Computed tomography (CT) scanning is a significant source of radiation. OBJECTIVE: To demonstrate that diagnostic quality CT images can be obtained while minimizing the effective radiation dose to the patient. METHODS: In this retrospective cross-sectional study, noncontrast head CT scan data were reviewed, and indications for scans and estimated radiation dose delivered were recorded. The estimated effective radiation dose (EERD) for each CT protocol was reviewed. RESULTS: We identified 251 head CT scans in a single month. Of these, 96 scans were using a low-dose shunt protocol with a mean EERD of 0.82 mSv. The remaining 155 scans were performed using the standard protocol, and the mean EERD was 1.65 mSv. Overall, the EERD was minimized while maintaining diagnostic scan quality. CONCLUSION: Although replacing a CT with magnetic resonance imaging is ideal to completely avoid ionizing radiation, this is not always practical or preferred. Therefore, it is important to have CT protocols in place that minimize radiation dose without sacrificing diagnostic quality. The protocols in place at our institution could be replicated at other academic and community hospitals and imaging centers. KEY WORDS: Child, CT protocol, Neuroimaging, Pediatrics, Radiation dosage, Tomography, X-ray computed Neurosurgery 77:228–232, 2015
DOI: 10.1227/NEU.0000000000000764
R
adiation exposure from medical imaging, especially computed tomography (CT) scanning, is of increasing concern. This is particularly true in the pediatric age group. Cancer risk due to CT scanning was highlighted in a recent publication by Pearce et al1 that reported an increased risk of brain cancer and leukemia in patients with higher lifetime exposure to CT radiation. Multiple methods have been proposed to limit the radiation exposure of children from diagnostic imaging. These techniques include implementation of protocols to limit the number of scans ordered and the replacement of CT scan with other modalities such as magnetic resonance imaging (MRI) and head ultrasound. There are situations, however, when CT is indicated, and in these situations, it is prudent to minimize the radiation dose applied. ABBREVIATION: EERD, estimated effective radiation dose
228 | VOLUME 77 | NUMBER 2 | AUGUST 2015
www.neurosurgery-online.com
Over many years, the neuroradiologists at our institution have developed protocols to minimize the estimated effective radiation dose (EERD) delivered with head CT scanning while maintaining diagnostic utility. The CT protocol selected for a particular patient depends on the indication for the scan and the age of the patient (Table 1). Here we present our age- and diagnosis-specific noncontrast head CT protocols. We evaluate the EERD administered to the patients stratified by indication for the scan and the age of the patient. Similar low-dose protocols could be applied anywhere CT scanning is performed, whether at a large academic referral hospital or at a smaller imaging center in the community.
METHODS For this retrospective cross-sectional study, the radiology records at our hospital were reviewed. We recorded data from all noncontrast head CT scans performed in a single month (October 2013). Volumetric scans performed for surgical planning
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
CT DOSE REDUCTION
TABLE 1. Noncontrast Head Computed Tomography Protocols Used at Our Institution Peak Killivoltage
Tube Current, mA
Time, s
80 100
50-100 50-100
0.75-1 0.75-1
100 100 120 120
80-200 80-230 80-230 80-300
0.75-1 0.75-1 0.75-1 0.75-1
Shunt ,1 y old .1 y old Standard ,2 y old 2-10 y old 10-18 y old .18 y old
(neuronavigation scans) were excluded. The dose length product was recorded during the acquisition of each study. The dose length product was then converted to the estimated effective radiation dose (EERD) in millisieverts using previously published age-based conversion factors (age 0-1 year = 0.011; age 1-5 years = 0.0067; age 5-10 years = 0.0040; age 10-15 years = 0.0032; age 15 and older years = 0.0021).2 Multiple head CT protocols are used at our institution depending on the indication for the scan and the age of the patient (Table 1). Patients for whom the indication was hydrocephalus, ventriculomegaly, or the presence of a cerebrospinal fluid (CSF) shunt were assigned to the “shunt” protocol group. All other patients were assigned to the “standard” protocol group. Because the purpose of this study was to evaluate how the scanning protocols are applied and the radiation received by patients, group assignments were based on the clinical indications for the scan rather than on the protocol used for the image acquisition. The majority of scans also used iterative reconstruction in postprocessing. All patients undergoing cranial imaging only have lead aprons applied to protect their thyroids and bodies. No special protection is applied to the corneas of the patients. The mean and standard deviation of the EERD was calculated for the shunt group patients and the standard group patients. Patients for whom the EERD was .2 standard deviations above the mean were considered outliers, and the individual scans reviewed to determine the reason for the increased radiation exposure. Patients were further stratified by age according to the head CT protocols at our institution, 2 subgroups for shunt protocol patients (younger than 1 year of age, 1 year of age and older) and 4 subgroups for standard protocol patients (younger than 2 years of age, 2-10 years of age, 10-18 years of age, 18 years of age and older). Again, the mean and standard deviation of the EERD were calculated for each subgroup. We randomly selected one-third of the scans for review by a pediatric neurosurgeon and a pediatric neuroradiologist. Each of these scans was scored on a 3-point scale (1 = ideal scan, 2 = not ideal but still diagnostic, 3 = nondiagnostic). All statistical calculations were performed with Microsoft Excel (Microsoft, Redmond, Washington) and StatPlus:mac (AnalystSoft, Inc, Alexandria, Virginia). This study was approved by our institution’s institutional review board.
RESULTS We identified a total of 251 noncontrast head CT scans performed in a single month. Patient ages ranged from 0 to 21
NEUROSURGERY
years. Of these, 96 scans were performed for hydrocephalus/ ventriculomegaly/CSF shunt and were designated as the shunt protocol. The remaining 155 scans were performed for a variety of indications including trauma, headache, macrocephaly, and seizure and were designated as the standard protocol (Figure). For the shunt protocol CT scans (Table 2), the mean EERD was 0.82 mSv with a standard deviation of 0.50 mSv (range, 0.24-3.02 mSv). Eighty-two of the scans (86%) were ,1.0 mSv. Six scans were .2 standard deviations above the mean. In 4 of these instances, the patient was scanned with a tube current that was greater than that specified for the shunt protocol. The other 2 patients were scanned according to protocol. Fifteen shunt protocol scans were performed in patients younger than 1 year of age. The mean EERD for this group was 1.01 mSv with a standard deviation of 0.45 mSv. The remaining 81 scans were in patients older than 1 year of age, and the mean EERD was 0.78 mSv with a standard deviation of 0.50 mSv. The standard protocol CT scans (Table 3) had a mean EERD of 1.65 mSv and a standard deviation of 0.97 mSv (range, 0.546.95 mSv). A total of 134 scans (87%) were ,2.5 mSv. Two scans delivered an EERD of .2 standard deviations above the mean. In both cases, the patient was scanned twice due to extreme motion artifact on the initial scan. Forty-nine standard protocol CT scans were performed in patients younger than 2 years of age. The mean EERD for this group was 2.52 mSv with a standard deviation of 1.25 mSv. Fortyfive patients were 2 to 10 years of age. The mean EERD for these patients was 1.31 mSv with a standard deviation of 0.42 mSv. Fifty-seven patients were 10 to 18 years of age and were exposed to a mean EERD of 1.24 mSv with a standard deviation of 0.38 mSv. The remaining 4 patients were 18 years of age or older. This group received a mean EERD of 0.97 mSv with a standard deviation of 0.16 mSv. We reviewed 84 of the scans for quality. Twenty-five of these scans were from the shunt protocol group and 59 of these scans were from the standard protocol group. This represents one-third of all the scans, 26% of the shunt protocol scans, and 38% of the standard protocol scans. No scan was given a score of 3 (nondiagnostic). Of the shunt protocol scans reviewed, 3 (12%) were scored as 2 (not ideal but still diagnostic). Two of these scans had motion artifact, and 1 did not use iterative reconstruction in postprocessing. Of the standard protocol scans, 2 (3%) were scored as 2. One of these was due to motion artifact and the other due to lack of iterative reconstruction in postprocessing. The Fisher exact test showed no statistically significant difference in the quality scores of the shunt protocol scans compared with the standard protocol scans (P = .15).
DISCUSSION The radiation dose used in CT scans has been of increasing concern in the past several years. There are now population-based clinical data demonstrating an increased cancer risk with increased lifetime exposure to diagnostic radiation,1 although a recently
VOLUME 77 | NUMBER 2 | AUGUST 2015 | 229
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
ALBERT AND GLASIER
FIGURE. Representative images obtained with the reduced radiation CT protocols. A, shunt, age 0.4 years, 0.65 mSv. B, shunt, age 4.8 years, 0.95 mSv. C, standard, age 0.4 years, 2.61 mSv. D, standard, age 5.0 years, 1.38 mSv. E, standard, age 12.1 years, 1.40 mSv. F, standard, age 18.9 years, 0.97 mSv.
published smaller clinical study suggests that this risk is low.3 Many groups have proposed methods of reducing radiation dose. These include replacing CT with MRI and performing low-dose CT scans with either a limited number of slices or reduced radiation dose per slice.
TABLE 2. Estimated Effective Radiation Dose Delivered for Scans Performed With Shunt Protocol (Hydrocephalus/Ventriculomegaly/ Cerebrospinal Fluid Shunt)a EERD, mSv
All studies Age ,1 y Age .1 y a
No.
Mean
SD
96 15 81
0.82 1.01 0.78
0.5 0.45 0.5
EERD, estimated effective radiation dose; SD, standard deviation.
230 | VOLUME 77 | NUMBER 2 | AUGUST 2015
Brain MRI Many groups have described replacing CT with MRI, particularly in nonemergent situations. This approach has been most widely described in patients with hydrocephalus and CSF shunts. Various sequences have been used including SSFSE, T1-FLASH, T2-HASTE, and diffusion-weighted imaging.4-7 We use rapid sequence MRI in our institution for routine hydrocephalus follow-up in cooperative patients. We routinely obtain images in all 3 anatomic planes with a scan time of 1 minute. We generally reserve this approach for patients whom we know can stay still without anesthesia. However, the primary limiting factor at our institution is the availability of MRI. During the daytime, the scanners are frequently occupied, and at night and on weekends, there is no in-house MRI technician. Although urgent scans are possible, there is a time delay that is not practical for emergency situations such as shunt malfunction and trauma. In addition, MRI lacks detail for bone imaging and visualization of shunt catheters. In our practice, it is
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
CT DOSE REDUCTION
TABLE 3. Estimated Effective Radiation Dose Delivered for Scans Performed With Standard Protocola EERD, mSv
All studies Age, y ,2 y 2-10 10-18 .18 a
n
Mean
SD
155
1.65
0.97
49 45 57 4
2.52 1.31 1.24 0.97
1.25 0.42 0.38 0.16
EERD, estimated effective radiation dose; SD, standard deviation.
still preferable or even necessary to use head CT scans for many patients. Limited-Slice CT Scan Two groups recently published studies exploring the use of limited-slice head CT scans as a means of reducing radiation dose in children with CSF shunts. Both studies used existing images from standard or reduced dose head CT scans and created theoretical limited-slice CT scans consisting of either 38 or 79 slices. Although appealing as a means to reduce radiation exposure, experience shows that some patients will have a subtle change in ventricle size that may be missed on limited-slice head CT. In addition, the cases of false-positive findings on limitedslice head CT could have led to unnecessary surgery. Also, limited-slice head CT scanning may not adequately visualize the ventricular catheter or overall ventricular anatomy, both of which are important for surgical planning. Finally, this is an approach that is only applicable to shunt follow-up and not suitable for trauma, altered mental status, macrocephaly, etc. Low-Dose Head CT Scan The most commonly reported approach for limited radiation exposure from medical imaging in children is the use of reduced-dose head CT scanning protocols. Various methods have been reported for decreasing the dose. Most commonly, this is accomplished by reducing the tube current (milliamperes) used for each study.10-12 Adjustments in tube rotation time have also been reported.10 In some cases, postprocessing is also used to reduce the noise in the image and therefore improve image quality.12 Reduced-dose CT scan protocols have also been of interest to craniofacial surgeons.13 Low-dose CT scanning is particularly applicable in these cases because lower radiation doses are needed for high-quality bone imaging than for brain imaging. In all these cases, significant reductions in effective radiation dose were achieved while maintaining the diagnostic utility of the imaging studies. Multiple CT protocols have been developed at our institution that optimize the radiation dose without sacrificing the diagnostic utility of the study (Table 1). Selection of the CT protocol
NEUROSURGERY
depends on both the indication for the scan and the age of the patient. Smaller patients are more susceptible to the effects of radiation, and reduced dose protocols in these children are important. The EERD for our standard protocol is lower than that for other widely used protocols. The shunt protocol radiation dose is even lower, 50% of the standard protocol. The dose reduction is most significant in the younger age groups. Regardless of the dose-reduction protocols, younger children still receive a larger EERD. This is expected given that younger children are more susceptible to the effects of radiation as indicated by the agedependent conversion factors. Overall, the reduced radiation protocols that we use have been successful in minimizing the effective radiation dose delivered to our patients while still allowing us to obtain diagnostic-quality images. As expected, a greater percentage of the shunt protocol scans were of lower quality. Generally, lower image quality is acceptable when the primary concern is evaluating ventricle size vs looking for potentially subtle changes in brain parenchyma. Therefore, there is a lower threshold for repeating the imaging of a standard protocol scan than of a shunt protocol scan. This repetition at the time of scanning is also why no scan in this study was scored as a 3 (nondiagnostic). Those patients with repeated scans were, however, exposed to greater levels of radiation, as noted earlier. We have also found that the use of iterative reconstruction is critical for the maintenance of diagnostic quality with these imaging protocols, as evidenced by the fact that several of the lower quality scans can be attributed to the lack of iterative reconstruction. CT scanners are more widely available than MRI scanners. Although it may be practical for large medical centers to use MRI instead of CT in most situations, this approach is not generalizable to all hospitals. Until such a time that all hospitals have ready access to MRI scanners around the clock, attention must be paid to limiting the radiation dose from CT scanning. The techniques that we use for radiation reduction could be easily applied at other institutions including those with limited access to MRI. Limitations This study is limited by its retrospective nature. In addition, specific information regarding the CT protocol to which individual patients were assigned was not reviewed. This allows for a realworld evaluation of our CT scanning protocols, as the radiation dose delivered to a particular patient may be higher than planned due to misassignment of the protocol, inappropriate scanner settings, or the need to repeat the scan for any reason. Finally, our dosimetry data indicate that our patients are being scanned at EERD lower than the pediatric reference standard levels. Therefore, we had no control group.
CONCLUSION The risk of ionizing radiation to children, particularly the increased risk of malignancy, is a valid concern. Head CT scanning
VOLUME 77 | NUMBER 2 | AUGUST 2015 | 231
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
ALBERT AND GLASIER
is a significant source of radiation for children. Multiple strategies exist for limiting the exposure of children to ionizing radiation from CT scans, including reducing the number of scans performed and replacing CT with MRI in selected instances. There are times that CT is necessary, and in these instances, it is important to minimize the radiation dose without sacrificing diagnostic yield. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Pearce MS, Salotti JA, Little MP, et al. Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet. 2012;380(9840):499-505. 2. Thomas KE, Wang B. Age-specific effective doses for pediatric MSCT examinations at a large children’s hospital using DLP conversion coefficients: a simple estimation method. Pediatr Radiol. 2008;38(6):645-656. 3. White IK, Shaikh KA, Moore RJ, et al. Risk of radiation-induced malignancies from CT scanning in children who underwent shunt treatment before 6 years of age: a retrospective cohort study with a minimum 10-year follow-up. J Neurosurg Pediatr. 2014;13(5):514-519. 4. O’Neill BR, Pruthi S, Bains H, et al. Rapid sequence magnetic resonance imaging in the assessment of children with hydrocephalus. World Neurosurg. 2013;80(6): e307-e312. 5. Wait SD, Lingo R, Boop FA, Einhaus SL. Eight-second MRI scan for evaluation of shunted hydrocephalus. Childs Nerv Syst. 2012;28(8):1237-1241. 6. Ashley WW Jr, McKinstry RC, Leonard JR, Smyth MD, Lee BC, Park TS. Use of rapid-sequence magnetic resonance imaging for evaluation of hydrocephalus in children. J Neurosurg. 2005;103(2 suppl):124-130. 7. Iskandar BJ, Sansone JM, Medow J, Rowley HA. The use of quick-brain magnetic resonance imaging in the evaluation of shunt-treated hydrocephalus. J Neurosurg. 2004;101(2 suppl):147-151. 8. Alhilali LM, Dohatcu AC, Fakhran S. Evaluation of a limited three-slice head CT protocol for monitoring patients with ventriculoperitoneal shunts. AJNR Am J Roentgenol. 2013;201(2):400-405. 9. Pindrik J, Huisman TA, Mahesh M, Tekes A, Ahn ES. Analysis of limitedsequence head computed tomography for children with shunted hydrocephalus: potential to reduce diagnostic radiation exposure. J Neurosurg Pediatr. 2013;12(5): 491-500. 10. Rybka K, Staniszewska AM, Biega nski T. Low-dose protocol for head CT in monitoring hydrocephalus in children. Med Sci Monit. 2007;13(suppl 1):147-151. 11. Udayasankar UK, Braithwaite K, Arvaniti M, et al. 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. 2008; 29(4):802-806. 12. Morton RP, Reynolds RM, Ramakrishna R, et al. Low-dose head computed tomography in children: a single institutional experience in pediatric radiation risk reduction: clinical article. J Neurosurg Pediatr. 2013;12(4):406-410. 13. Vazquez JL, Pombar MA, Pumar JM, del Campo VM. Optimised low-dose multidetector CT protocol for children with cranial deformity. Eur Radiol. 2013; 23(8):2279-2287.
Acknowledgments The authors thank Donna Hoover, BSRT(R)(MR)(CT), for helping to identify the patients and scans for this study and Daniel Harton, BSRT(R)(CT), for providing the CT protocols.
232 | VOLUME 77 | NUMBER 2 | AUGUST 2015
COMMENTS
A
lthough this was a retrospective, single-center study, the authors accumulated a large number of patients over a short time period and managed to demonstrate their premise that diagnostic quality scans can be achieved with reduced radiation dose when CT must be used. I was particularly pleased to see images taken under the lower dose protocol and the standard, demonstrating the preserved quality of the images and the diagnostic effectiveness. Neurosurgeons, especially pediatric neurosurgeons, will want to have this information to assist with patient care. Attention must be given to the details of the technique used by this center to achieve dose reduction with useful scan quality. As they discuss, other strategies used for a similar goal have limitations. Ann Marie Flannery Lafayette, Louisiana
T
he authors document a significant reduction in the estimated effective radiation dose (EERD) for children undergoing CT imaging of the brain using a CSF shunt protocol. The authors only specify protocol-based limitations in peak kilovoltage and tube current, based on age. Despite the noted reductions in EERD, no shunt protocol scan was nondiagnostic, although there was a trend toward more frequent occurrence of “not ideal but still diagnostic” scans in this group (12% for shunt protocol vs 3% for standard protocol scans). In real-life clinical practice, using an “exaggerated sniff” patient position in the CT scanner can achieve substantial additional reduction in EERD, while largely avoiding exposure of the thyroid and corneas to radiation.1 The ideal strategy for radiation exposure reduction during pediatric axial brain imaging is to substitute MRI for CT imaging.2 As the authors point out, however, MRI for emergency indications is not logistically or financially feasible in many centers. In addition, simple strategies are also important: judicious ordering of axial imaging appropriate to the clinical situation, use of radiography rather than CT when clinically appropriate (for example, in most cases of cervical spine clearance after minor trauma), and restricting the use of repeat imaging in trauma for nonevolving pathologies.3 The authors have contributed valuable information to an ongoing effort to reduce the exposure of pediatric patients to CT radiation. This is particularly important for young patients, such as those with hydrocephalus, often exposed to frequent axial brain imaging. Nathan R. Selden Portland, Oregon
1. Didier RA, Kuang AA, Schwartz DL, Selden NR, Stevens DM, Bardo ME. Decreasing the effective radiation dose in pediatric craniofacial CT by changing head position. Pediatr Radiol. 2010;40(12):1910-1917. 2. Yue EL, Meckler GD, Fleischman RJ, et al. Test characteristics of quick brain MRI for shunt evaluation in children: an alternative modality to avoid radiation. J Neurosurg Pediatr. Epub 2015 Jan 30. doi: 10.3171/2014.9.PEDS14207. 3. Durham SR, Liu KC, Selden NR. Utility of serial computed tomography imaging in pediatric head trauma. J Neurosurg. 2006;105(5 suppl):365-369.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
