8/18/2011 Acknowledgements Managing Pediatric CT Patient Doses Marilyn Goske, MD John Boone, PhD Cynthia McCollough, PhD Michael McNitt-Grey, PhD Dianna Cody, PhD Tom Toth, PhD Mahesh Mahadevappa, PhD G. Donald Frey, PhD Keith J. Strauss, MSc, FAAPM, FACR President X-Ray Computations, Inc. Boston, Massachusetts MISSION STATEMENT INTRODUCTION A. B. C. D. E. F. G. H. Image Gently Campaign Pediatric CT Public Health Concerns Shortcomings of Dose Indices for CT Affect of Patient Size on Dose Indices Clinical Dilemma Interim Solution: AAPM TG204 Applications of SSDE Managing Pediatric CT Patient Doses Alliance for Radiation Safety in Pediatric Imaging is a coalition of health care organizations dedicated to providing safe, high quality pediatric imaging worldwide. The primary objective is to raise awareness in the imaging community of the need to adjust radiation dose when imaging children. Adapted from Goske 1 8/18/2011 Does Practice Need to Change? How much do we really understand? Image Gently A. The ultimate goal of the Alliance is to accelerate the change of local practice. 1. 2. Problem? Scientific observation to local practice change ~ 17 years!1 1Greenberg Under estimation by 75% of MDs! SB. Trans Clin Climatol Assoc 119:2450261, 2008 Lee et al. Radiology. 2004; 231:393-398 Adapted from Goske Does Practice Need to Change? How much do we really understand? Does Practice Need to Change? How much do we really understand? Which of the following provides a reasonable estimate of a child‟s CT dose? Which of the following provides a reasonable estimate of a child‟s CT dose? A. B. C. D. A. B. C. D. CTDIvol (mGy) DLP (mGy-cm) Effective Dose (mSv) Values in DICOM Structured Report E. None of the above! CTDIvol (mGy) DLP (mGy-cm) Effective Dose (mSv) Values in DICOM Structured Report E. All of the above! 2 8/18/2011 Critical Triad Founding Organizations The Society for Pediatric Radiology To act together for radiation protection for children American Society of Radiologic Technologists American Association of Physicists in Medicine American College of Radiology Medical physicists Adapted from Goske Adapted from Goske What is Image Gently An education, awareness and advocacy campaign To improve radiation protection for children worldwide Alliance for Radiation Safety in Pediatric Imaging 58 health care organizations/agencies 800,000 radiologists, radiology technologists, medical physicists worldwide Adapted from Goske Objectives Decrease radiation dose in children Positive message ..resulting dose to population will lead to higher cancer rates, accounting for as many as 2% of all cancers in the U.S. Enroll key organizations Increase awareness educate advocate change practice Adapted from Goske 3 8/18/2011 Use internal communication of each member organization Campaigns in CT CR / DR Interventional Radiology Flouroscopy Nuclear medicine Imagegently.org Adapted from Goske Adapted from Goske PEDIATRIC CONSIDERATIONS CLINICAL EDUCATIONAL MATERIALS Abdomen Baseline: PA Thickness (cm) 9 12 14 16 19 22 25 31 kVp fill in Approx Age newborn 1 yr 5 yr 10 yr 15 yr small adult med adult large adult mA Time (sec) fill in fill in Abdomen mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.43 #VALUE! 0.51 #VALUE! 0.59 #VALUE! 0.66 #VALUE! 0.76 #VALUE! 0.90 #VALUE! 1.0 fill in 1.27 #VALUE! Abdomen Baseline: PA Thickness (cm) 9 12 14 16 19 22 25 31 Pitch Abdomen Pitch Thorax fill in fill in Thorax mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.42 #VALUE! 0.49 #VALUE! 0.57 #VALUE! 0.64 #VALUE! 0.73 #VALUE! 0.82 #VALUE! 0.91 #VALUE! 1.16 #VALUE! kVp 120 Approx Age newborn 1 yr 5 yr 10 yr 15 yr small adult med adult large adult IMAGE GENTLY BODY mA Time (sec) 400 1 Abdomen mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.43 172 0.51 204 0.59 236 0.66 264 0.76 304 0.90 360 1.0 400 1.27 508 Pitch Abdomen Pitch Thorax 1.25 1.5 Thorax mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.42 202 0.49 235 0.57 274 0.64 307 0.73 350 0.82 394 0.91 437 1.16 557 PEDIATRIC CONSIDERATIONS CLINICAL EDUCATIONAL MATERIALS 1. CTDIvol for Adults a. < 25 mGy Body b. < 75 mGy Head 2. Pediatric Patient Dose < Adult Dose 3. Conservatively high 4. Developed for adult department that images children occasionally 4 8/18/2011 Image Gently Vendor Summit 8 FREE online modules for CT technologists August 20, 2008 CTDIvol is not an indication of patient dose Pediatric Radiologists and Technologists have been asking for a reasonable estimate of CT radiation dose in children for over 10 years. You need to provide better training on pediatric applications of your images! Toshiba, GE Healthcare, Philips, Siemens Adapted from Goske TRAINING Your Lender Training is only as effective as the trainer‟s understanding of their trainees Each trainee may have a bit different Your Buyer Perspective! Pediatric needs are Different Your Appraiser YOUR HOUSE as seen by... Yourself Your Tax Assessor 5 8/18/2011 Preliminary estimate of changes in Medical radiation exposure to US population US 1980* Procedures vs Effective dose contributions Interventional 2% US 2006 Other 0.06 CT 12% Nuclear Medicine 3% Radiography & Fluoroscopy* 19% Interventional 12% CT 46% Interventional 0.4 mSv Radiography 0.6 mSv Medical 0.54 Nuclear Medicine 0.7 mSv Natural 3.0 mSv Natural ?? (3.0 mSv) CT 1.5 mSv Radiography & Fluoroscopy* 83% Nuclear Medicine 23% Percent Procedures Medical 0.54 mSv per capita Total 3.6 mSv per capita Medical 3.2 mSv per capita Total 6.2 mSv per capita Effective Dose Contributions 17% of All Exams Deliver 81% of Total Effective dose 91% of Pediatric Dose in the ED comes from CT * NCRP 93 Adapted from Mahesh Adapted from Mahesh GROWTH OF MEDICALRADIATION DOSE C. General Radiography 1. 85% of dose from exams Pelvis/Hips of Abdomen Upper GI Pelvis/Hips, Spine Abdomen or Spine Head Face Chest 2. 280,000 IVP person Sv Barium enema Extremities 3. ~ 0.62 mSv Mammography Adapted from Mahesh Number millions % Collective Person Sv % 20 5 45,700 24.8 4 1 38,000 20.7 17 4 57,000 31.0 15 4 16,000 8.7 2.4 0.6 2,300 1.2 126 31 6,500 3.5 1.2 0.3 3,700 2.0 0.65 0.2 8,500 4.6 56 14 1,900 1.0 34 9 2,200 1.2 Dental 125 31 2,300 1.2 Total ~401 PEDIATRIC CONSIDERATIONS A. Radiation Induced Cancer Lifetime Risk From 1 Sv Dose 1. Average a. 5% Males b. 6% Females 2. First Decade 13 - 15% 3. Middle Age 2-3% 4. Children 3 – 5 times more sensitive Adapted from Hall ~184,000 6 8/18/2011 PEDIATRIC CONSIDERATIONS Radiation Risk based on Effective Dose? A. Almen and S. Mattsson, J. Radiol. Prot. 16 (2), 81-89 (1996). Somatic risk %/Sv Children Aged 0-9 Hereditary risk %/Sv 14.5 Total risk %/Sv 2.5 Real World …. Radiation distribution crosses the imaged volume Peak dose 17 “Tails” of dose distribution Children Aged 10-19 8.5 2.5 11 Whole population 5 1 6 - 5 mm 0 + 5 mm Adapted from Frey CTDI = Integral under the radiation dose profile along the z-axis from a single axial scan of width nT. radiation dose profile nT “nominal beam width” = “total nominal scan width” Adapted from Frey Z CT SCANNER DOSE INDICES B. Measurement of CT Radiation Dose 1. Plastic cylindrical phantoms: CTDI Phantoms a. (PMMA) b. 16 & 32 cm diameter 2. Pencil chamber moved into provided holes to measure radiation dose a. 1 cm under peripheral surface PMMA hole b. Center of phantom plug center c. Non measured 100 mm hole holes plugged pencil Adapted from TG204 chamber 7 8/18/2011 CTDIvol = [2 * surface dose / 3 + central dose / 3 ] / Pitch CT SCANNER DOSE INDICES C. Measure CTDIvol 1. Measure CTDIvol with identical scan parameters a. b. c. d. Measured CTDIvol = 47 Measured CTDIvol = 37 21.6 mGy 38 mGy 47 mGy 47 mGy 35 mGy kVp mA Rotation time Bow Tie Filter 2. Use phantom 10, 16, and 32 cm diameter Measured CTDIvol = 18 10.8 mGy 10 cm Diameter 16 cm Diameter 32 cm Diameter Measured CTDIvol increases 2.6 times as phantom size decreases! 8 8/18/2011 CT SCANNER DOSE INDICES D. Displayed CTDIvol 1. Dose that represents distribution of dose given to cross-sectional area of a slab of the CTDI phantom (16 or 32 cm diameter) 2. Reflects changes in: a. b. c. d. e. f. High voltage to x-ray tube (kVp) X-ray tube current (mA) Rotation time (sec) Bow tie filter shape, thickness, material Pitch Source to detector distance CT SCANNER DOSE INDICES D. Displayed CTDIvol 5. does not represent . . . Patient dose!! CT SCANNER DOSE INDICES D. Displayed CTDIvol 3. Standardized method to estimate and compare the radiation output of two different CT scanners to same phantom. 4. Dose index of CT scanners if the fan beam width in z direction of the patient is small (< 1 cm) a. AAPM TG111 addresses these shortcomings b. Beyond scope of this presentation. CT SCANNER DOSE INDICES E. Displayed Dose Length Product (DLP) 1. DLP (mGycm) = CTDIvol * Scan Length a. Scan length is the length of phantom irradiated. b. „Represents‟ energy transferred. 2. DLP is not a patient dose index because CTDIvol does not represent patient dose. 9 8/18/2011 Adapted from Frey DLP = 200 mGy•cm CTDIvol = 20 mGy ten 1-cm slices CT SCANNER DOSE INDICES E. Displayed Dose Length Product (DLP) DLP = 400 mGy•cm CTDIvol STILL = 20 mGy twenty 1-cm slices So, DLP represents the greater biologic risk! AFFECT of PATIENT SIZE on DISPLAYED CTDIVOL & DLP Large Adult Adult 5 year 1. Previous example: radiation doses equal. 2. Second patient at greater risk because a larger length of body was scanned. 3. While DLP is the dose to a standard phantom, increases in DLP indicate increases in risk to patient. AFFECT of PATIENT SIZE on DISPLAYED CTDIVOL & DLP A. Patients of a given age vary dramatically in size. Abdomens of the largest 3 year olds are approximately the same size as the abdomens of the smallest adults. 1 year Neonate 4 cm Size of patient, not age should be used. 1 HVL @ 120 KVP ~ 70 keV 10 8/18/2011 AFFECT of PATIENT SIZE on DISPLAYED CTDIVOL & DLP Measured CTDIvol = 47 Measured CTDIvol = 37 Measured CTDIvol = 18 21.6 mGy 38 mGy 47 mGy 35 mGy 47 mGy 10.8 mGy AFFECT of PATIENT SIZE on DISPLAYED CTDIVOL & DLP CLINICAL DILEMMA A. Displayed CTDIvol is independent of the patient size; displayed CTDIvol assumes either 16 or 32 cm CTDI phantom. B. 16 cm CTDI phantom: adult dose over while pediatric dose under estimated. C. 32 cm CTDI phantom: adult and pediatric dose under estimated ~ 2.5 times! Displayed CTDIvol16 = 37 Displayed CTDIvol16 = 37 Displayed CTDIvol16 = 37 Displayed CTDIvol32 = 18 Displayed CTDIvol32 = 18 Displayed CTDIvol32 = 18 D. Propagated by DICOM Structured Reports and CT scanner dose reports. 11 8/18/2011 TG 204 ! ! "# $%&' ( )*$+ ( ,$- . / $ ! E. Report does not: ! ! ! 1. Address correction factors for heads ! "#$%!"&%' #(#' !) *+%!, +-#. / -%+!0"") , 1!#2!3%4#/ -5#' ! / 24!6478-!9*4: !; <!, =/ . #2/ -#*2+! ! ! ! ! %&' ( )*$( 0$! ! "# $1234$5 )( 6' $- . / 7$89$: ( ;;2<( )2*8( 9$= 8*>$*>&$?9*&)92*8( 92;$ @( A A 8338( 9$( 9$%2B8( ;( C8: 2;$D98*3$29B$# &236)&A &9*3$E?@%DF$29B$*>&$ ?A 2C&$5 &9*;G$: 2A ' 28C9$( 0$*>&$! ;;829: &$0( )$%2B82*8( 9$H20&*G$89$"&B82*)8: $ ?A 2C89C,$ ! ! ! ! ! ! ! ! ! ! 2. Correct small (< 1%) doses from scanned projection images 3. Correct for variation (~ 5%) in attenuation of thorax vs abdomen 4. Correct small variation in pre and post contrast scans TG 204 TG 204 E. Report does not address: 5. Changes in dose as a function of fan beam a. 20 cm: relative dose ~ 2.7 b. 30 cm: relative dose ~ 3 c. Dose increases ~ 10% d. Dose Profiles vs Fan Beam Width F. Data from four independent investigators studying patient size correction factors. Adapted from TG 204 1. Physical measurements on phantoms A. Anthropomorphic Phantoms (McCollough Laboratory “Mc”) B. Cylindrical PMMA phantoms (Toth / Strauss Collaboration “TS”) 2. Monte Carlo computer modeling 30 cm 20 cm Adapted from J. Boone C. Monte Carlo Voxelized Phantoms (McNitt-Gray Laboratory “MG”) D. Monte Carlo Mathematical Cylinders (Boone Laboratory “Z-B”) 12 8/18/2011 TG 204 TG 204 age in years 0 1 5 1 0 1 5 32 cm 120 kVp Adapted from TG 204 TG 204 1 5 1 0 1 5 Adapted from TG 204 TG 204 G. What about scans performed at 80, 100, or 140 kVp? ≥1 Year 16 cm 120 kVp age in years 0 1. 5% difference overall 2. 3% difference between 1 yr old (15 cm) & adult (32 cm) H. What about scans performed in the thorax? 1. Thorax data from Huda et al. 2. 16% dif @ 12 cm 3. 7% dif @ 17 cm 4. < 3% dif > 17 cm from 120 kVp only Combined TS / ZB: 80-140 kV Adapted from TG 204 Adapted from Boone 13 8/18/2011 TG 204 TG 204 I. What is an effective diameter? 1. Circle with area of patient’s cross section 2. Effective diameter can be estimated if the patient’s AP or lateral dimension is known. J. Fitted equations calculate the effective diameter from either AP or LAT dimension of patient based on published data of actual patient sizes. Adapted from TG 204 Adapted from TG 204 circle of equal area ICRU 74 Strauss Boone Fit to All ICRU 74 Strauss Boone Fit to All AP effective diameter lateral 204 K. What if I am TG doing retrospective dose analysis and I only know age of patient? 1. Corrections based on patient size are more accurate. TG 204 AP or PA Projection Scan Adapted from TG204 L. Determining patient size 1. Measure Lateral dimension with mechanical calipers. 2. Measure Lateral or AP dimension from AP or Lateral projection scan. 3. Measure AP or LAT dimension from axial scan view. 4. In #2 & #3 patient must be centered in gantry. a. Magnification Error Adapted from TG 204 AP or PA Projection Scan Adapted from TG204 Effective Diameter as a function of age per ICRU 74 14 8/18/2011 1. Failure to identify correct phantom, 16 or 32 cm leads to a systematic error of 100%. 2. No standard exists. Choice may depend on: a. b. c. d. Selected protocol: adult or pediatric Selected scan field of view Year of manufacture Software level 3. Make no assumptions: contact manufacturer of your unit through its service organization. TG 204 O. Correction Factor based on 32 cm CTDI Phantom TG 204 M. Determining size of CTDI phantom your CT scanner used to estimate CTDIvol Adapted from TG 204 TG 204 TG 204 Q. SSDE Accuracy 1. 20% Correction Factor based on 16 cm CTDI Phantom P. 2. Product is an estimate of patient dose 3. Report dose estimates with proper number of significant digits a. SSDE > 5 mGy: integers only, e.g. 7 or 23 mGy b. SSDE < 5 mGy: one decimal point, e.g. 2.7 or 4.5 mGy Adapted from TG 204 15 8/18/2011 TG 204 R. Dose Reporting by Radiologists The CTDIvol value reported on the scanner for the [32 or 16] PMMA phantom was used with correction factors obtained from AAPM Report 204. The correction factor for this patient was based on the patient’s [AP, LAT, AP + LAT, or effective dimension] This method is thought to produce dose estimates with accuracy to within 20%. For this patient, the size corrected (SSDE) estimate for this CT scan is _____ mGy. SAMPLE CALCULATION: PRESCAN A. Determine size of patient 1. AP Projection Scan: 16.8 cm B. 16 cm CTDI phantom used by scanner to calculate CTDIvol C. Displayed CTDIvol = 9.29 mGy D. 9.29 mGy x 1.08 = 10 mGy SSDE Adapted from TG 204 Adapted from TG 204 SAMPLE CALCULATION: POST SCAN SAMPLE CALCULATION: POST SCAN A. Determine size of patient 1. AP = 9.9 cm; LAT = 12.3 cm 2. AP + LAT = 22.2 cm A. Determine size of patient 1. AP = 9.9 cm; LAT = 12.3 cm 2. AP + LAT = 22.2 cm B. 32 cm CTDI phantom assumed B. 16 cm CTDI phantom assumed C. Displayed CTDIvol = 5.4 mGy C. Displayed CTDIvol = 10.8 mGy D. 5.4 mGy x 2.5 = 13 mGy SSDE D. 10.8 mGy x 1.24 = 13 mGy SSDE Adapted from TG 204 Adapted from TG 204 16 8/18/2011 CLINICAL CHALLENGE A. Child imaged on GE CT on Tuesday followed by second examination on Wednesday on Siemens CT in same department. 1. GE: Displayed CTDIvol(16) = 10.8 mGy 2. Siemens: Displayed CTDIvol(32) = 5.4 mGy CLINICAL CHALLENGE D. Correction Factors for Heads 1. Will be published in the future 2. Correction factor not as large 32 cm CTDI 16 cm CTDI B. Mom & Dad were not happy campers! C. SSDE = 13 mGy for both studies! Applications of SSDE A. Understand difference between SSDE and CTDIvol 1. SSDE may be useful as a first approximation of average organ dose to organs completely contained in scan volume. 2. SSDE can NOT be substituted in place of CTDIvol when using k-factors to estimate Effective Doses from CT exam Applications of SSDE B. Start calculating SSDEs 1. More accurate estimates of patient dose than CTDIvol 2. Recorded values of SSDE in patient medical record have more applications than CTDIvol. 3. DRLs could be based on SSDE values 4. Recorded SSDE values will lead to department dose standards. 17 8/18/2011 Applications of SSDE B. Start calculating SSDEs 5. SSDE values can be used as a tool to manage patient doses from CT. 6. Work ongoing to get CT manufacturers to calculate and display SSDE values. Managing Pediatric CT Patient Doses C. Establish Appropriate Pediatric Dose Levels by reducing mAs appropriately 1. Reduce mAs until pediatric SSDE < adult SSDE for department. a. 10 cm AP b. 1 / 2.5 = 40% original mAs Managing Pediatric CT Patient Doses A. CTDIvol Measurements 1. 32 cm adult body measurement 2. 16 cm adult head measurement B. Compare/adjust wrt ACR Reference Values Establish reasonable adult doses Are they high, moderate, or low? 1. < 25 mGy 2. < 75 mGy Managing Pediatric CT Patient Doses C. Establish Appropriate Pediatric Dose Levels by reducing mAs appropriately 1. Reduce mAs until pediatric SSDE < adult SSDE for department. 2. IG website model of mAs reduction a. 10 cm AP dimension b. ~ 45% original mAs 18 8/18/2011 PEDIATRIC CONSIDERATIONS CLINICAL EDUCATIONAL MATERIALS Abdomen Baseline: PA Thickness (cm) 9 12 14 16 19 22 25 31 kVp fill in Approx Age newborn 1 yr 5 yr 10 yr 15 yr small adult med adult large adult mA fill in Time (sec) fill in Abdomen mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.43 #VALUE! 0.51 #VALUE! 0.59 #VALUE! 0.66 #VALUE! 0.76 #VALUE! 0.90 #VALUE! 1.0 fill in 1.27 #VALUE! Abdomen Baseline: PA Thickness (cm) 9 12 14 16 19 22 25 31 Pitch Abdomen Pitch Thorax fill in fill in Thorax mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.42 #VALUE! 0.49 #VALUE! 0.57 #VALUE! 0.64 #VALUE! 0.73 #VALUE! 0.82 #VALUE! 0.91 #VALUE! 1.16 #VALUE! kVp 120 Approx Age newborn 1 yr 5 yr 10 yr 15 yr small adult med adult large adult mA 400 Managing Pediatric CT Patient Doses D. Dose reduction of IG website results in conservatively high doses IMAGE GENTLY BODY Time (sec) 1 Abdomen mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.43 172 0.51 204 0.59 236 0.66 264 0.76 304 0.90 360 1.0 400 1.27 508 Pitch Abdomen Pitch Thorax 1.25 1.5 Thorax mAs Reduction Estimated mAs = Factor (RF) BL x RF 0.42 202 0.49 235 0.57 274 0.64 307 0.73 350 0.82 394 0.91 437 1.16 557 Managing Pediatric CT Patient Doses F. Should voltages < 120 kVp be used for Children? 1. Reduced high voltage at same dose a. Dial up reduced mAs technique b. Note displayed CTDIvol 120 c. Reduce kVp d. mAs up until CTDIvol 80 = CTDIvol 120 e. Increased Contrast at same dose E. Doses in a pediatric department might be: 1. 2. 3. 4. 5. 10 cm (Newborn) 11 cm (1 yr old) 14 cm (5 yr old) 17 cm (15 yr old) 23 cm (Adult) ~ 0.6 x adult SSDE ~ 0.7 x adult SSDE ~ 0.8 x adult SSDE ~ 0.9 x adult SSDE adult SSDE Managing Pediatric CT Patient Doses 2. Reduced high voltage; reduced dose a. Dial up reduced mAs technique b. Note displayed CTDIvol 120 c. Measure increased contrast at 80 kVp compared to 120 kVp. i. ACR accreditation phantom ii. Clinical FoV / Bow tie Filter 19 8/18/2011 Managing Pediatric CT Patient Doses 2. Reduced high voltage; reduced dose d. Estimate increase in noise by comparing CTDIvol 120 & CTDIvol 80 e. Contrast Up 40% / Noise Up 60% f. Increase mAs at 80 kVp until CNR80 kVp = CNR 120 kVp g. Same Image Quality; Reduced Dose Conclusions A. Adult hospitals performing 80% of all pediatric patients should manage pediatric radiation doses. 1. Use adult protocols and calculate adult SSDE. 2. SSDE of pediatric patients prior to scan < SSDE for adult patient. 3. Changing kVp in addition to mAs is more involved, but allows both image quality improvement and/or dose reduction. 20