1. Computed Tomography (CT) Dose Optimization
Petrone Associates, LLC
Specialists in Applied Medical Physics

2. CT Dose in the News
Americans' Radiation Exposure Rises 6-Fold in 29 Years: CT Scans and Other Radiation-Based Medical Tests May Be to Blame for Increase in Radiation Exposure
Audrey Grayson
Study Finds Increase In Diagnostic Tests Ordered
CBS 2: Los Angeles
Overexposed: Imaging tests boost U.S. radiation dose
Julie Steenhuysen
New England Journal of Medicine Article – Nov. 2007
Dr.s. David Brenner and Eric Hall

3. Rise in Medical Radiation Exposure
In 2006, 62 million CT scans performed in U.S.
NCRP Report No. 160, "Ionizing Radiation Exposure of the Population of the U.S.“ (2006)
Americans exposed to 7 times amount of radiation in 1980s
CT contribution:
24% of total radiation exposure
50% of medical radiation exposure
In 2006, approximately 62 millions CT scans were performed in the United States, an increase from the 26 million performed in 1996 and approximately 3 million in 1980.

4. Computed tomography (medical) (24%)
2006 U.S. Radiation Exposure: Collective Effective Dose (%) NCRP Report No. 160, "Ionizing Radiation Exposure of the Population of the United States"
Space (3%)
(background)
Internal
(background)
(3%)
Terrestrial
(background)
(3%)
Radon & Thoron
(background) (37%)
Computed tomography (medical) (24%)
In 2006, CT was estimated to contribute 24% of the total population radiation exposure in the United States and 50% of all medical exposures.
Industrial (<0.1%)
Occupational (<0.1%)
Consumer (2%)
Conventional radiography & fluoroscopy (medical)
(5%)
Nuclear Medicine (medical) (12%)
Interventional fluoro (medical)
(7%)

5. Increased Use of CT Helical & Spiral CT
Dual-slice & Multi-Detector CT (MDCT)
Diagnostic medical reasons for use increase:
256+ slice scanners
Decreased scan time = more CT examinations
Better images lead to increased clinical use (vascular & cardiac exams, virtual colonoscopy, etc.)
Multi-planar Reconstruction (MPR)
Non-medical reasons for increased use:
Overcautious ordering of CT series for fear of malpractice litigation
Financial incentive due to fee-for-service reimbursement
Justify cause of newly purchases equipment/technology upgrades

6. CT Dose Contributors
Radiation dose in CT is unique in comparison to general radiography:
Highly collimated beam = smaller volume irradiated by primary beam
Volume of tissue exposed from almost all angles during rotational acquisition
Higher techniques (kVp & mAs) to achieve higher signal-to-noise ratio (i.e. better contrast resolution)
Considerable amount of Scattered dose

7. CT Dose Contributors Dose Factors:
- tube voltage (kVp), tube current (mA), scan time, patient size, scan length, pitch, beam width, type of scanner
Multiple Series in Scans of Abdomen & Pelvis
- Pre-contrast, Post-contrast
at least 3: 30% of patients
5+ scans: 7%
9 or more: 4%

8. Publications / Regulations
Journal of American College of Radiology
Special Issue: Radiation Dose Optimization (2014)
The Joint Commission
Effective July 1, 2015: NEW Diagnostic Imaging Standards
Includes CT technologist participation in ongoing training and dose optimization

9. Reducing CT Dose ACR Appropriateness Criteria® (AC) MRI / Ultrasound
evidence-based guidelines to assist referring physicians in making the most appropriate imaging or treatment decision for a specific clinical condition
MRI / Ultrasound
Perform non-ionizing radiation exams when possible
Limit the number of CT Scan series
Reduce Pre- and Post-Contrast study orders when only a single series will suffice
Use Multiple Recons for the same series
The largest reduction in CT dose would be the reduction of unnecessary exams

10. CT Dose Descriptors CTDIW = 1/3 center CTDI100 + 2/3 edge CTDI100
CTDI: Computed Tomography Dose Index
CTDI100: Measured by physicist with 100 mm ionization chamber and cylindrical phantom
16 cm phantom: represents Adult Head, Pediatric Head & Pediatric Abdomen
32 cm phantom: Adult Abdomen
CTDIW = 1/3 center CTDI /3 edge CTDI100
CTDIVOL = CTDIW / pitch (This is what is indicated on the CT console)
Dose Length Product (DLP) = CTDIVOL x Scan Length (cm)
Dose Length Product is the product of the length of the irradiated scan volume and the average CTDIvol over that distance

11. Defining CT Dose
CTDIVOL is a standardized parameter used to measure scanner radiation output
CTDIVOL is NOT a patient dose
CTDIVOL (mGy) is reported for either a 16 cm or 32 cm reference phantom
For the same technique (eg. kVp, mAs, etc.) the CTDIVOL for the 16 cm phantom is about twice that of the 32 cm phantom

12. How is CTDIVOL related to patient dose?
CTDIVOL is NOT a patient dose
The relationship between the two depends on many factors including patient size and composition
AAPM Report 204 introduces a parameter:
Size Specific Dose Estimate (SSDE)
For the same CTDIVOL, a smaller patient will tend to have a higher patient dose than a larger patient
The Size Specific Dose Estimate allows for the estimation of patient dose based on CTDIvol and patient size. For the same CTDIVOL, a smaller patient will tend to have a higher patient dose than a larger patient because the dose is distributed over a smaller volume of tissue

13. How is CTDIVOL related to patient dose?
120 kVp at 200 mAs
120 kVp at 200 mAs
32 cm Phantom
CTDIVOL = 20 mGy
CTDIVOL = 20 mGy
Both patients scanned with the same CTDIVOL Patient dose will be higher for the smaller patient

14. How is CTDIVOL related to patient dose?
120 kVp at 100 mAs
120 kVp at 200 mAs
32 cm Phantom
CTDIVOL = 10 mGy
CTDIVOL = 20 mGy
Smaller patient scanned with a lower CTDIVOL Patient doses will be approximately equal

15. Size Specific Dose Estimate (SSDE)
AAPM report 204 describes a method to calculate SSDE using CTDIVOL
Conversion factors based on patient size (e.g., AP or lateral width, effective diameter) are provided to estimate patient dose for a patient of that size

16. How is CTDIVOL related to patient dose?
120 kVp at 100 mAs
120 kVp at 200 mAs
32 cm Phantom
9 cm
27 cm
CTDIVOL = 10 mGy
SSDE = 13.2 mGy
CTDIVOL = 20 mGy
SSDE = 13.2 mGy
Patients have equivalent SSDE

17. So Why Use CTDIVOL?
CTDIVOL provides information about the amount of radiation used to perform the study
CTDIVOL is a useful index to track across patients and protocols for quality assurance purposes
CTDIVOL can be used as a metric to compare protocols across different practices and scanners when related variables, such as resultant image quality, are also taken in account
The ACR Dose Index Registry (DIR) allows comparison across institutions of CTDIVOL for similar exam types (e.g. routine head exam)

18. CT Acquisition Parameters/Settings
Define the technique that will be used and how the scan will proceed
Are set in the user interface where scans are prescribed
Changing a single Acquisition Parameter while holding everything else constant will typically affect the CTDIVOL for that scan
CT Scan Parameters:
Scan mode
Detector Configuration
Table Speed/Increment
Pitch
Tube Rotation Time
Tube Current (mA)
Tube Potential (kVp)
Field of View
Filtration
The relationships between acquisition parameters and CTDIvol described in the following slides assume all other parameters are held constant

19. Scan Modes
Each scan mode differ in how the table/patient moves during the scan
Axial
table is incremented between slice acquisitions
Helical / Spiral
table moves continuously during acquisition
Dynamic
Table is stationary for multiple slice acquisitions in the same area
Example: Bolus tracking, Brain Perfusion
Dynamic scans have a larger CTDIVOL due to the summation of the CTDIVOL values for each rotation

20. Detector Configuration
Is the combination of the number of data channels (N) and the width or thickness (T) of the detector associated with each data channel
The Detector Configuration determines the Beam Width or Beam Collimation (N x T)
For a selected detector width per data channel, a smaller total Beam Collimation usually has a higher CTDIVOL than a larger Beam Collimation
Example: For a 16 slice scanner with a detector width per channel of 1.25 mm, a collimation of 4 x 1.25 mm is generally LESS DOSE EFFICIENT than a collimation of 16 x 1.25 mm

21. Detector Configuration

22. Detector Configuration
For a selected detector width (T), a smaller total Beam Collimation (N x T) usually has a higher CTDIVOL than a larger Beam Collimation
Example: For a 16 slice scanner with a detector width per channel of 1.25 mm, a collimation of 4 x 1.25 mm is generally less dose efficient than a collimation of 16 x 1.25 mm
Because the actual x-ray beam is slightly wider than the beam width (N x T), there is overlap which increases the radiation dose to the patient

23. Table Speed & Pitch
Table speed (aka feed or increment) is the movement of the table through the bore of the scanner over a full 360 degree rotation
Units: millimeters/rotation or millimeters/second
PITCH is the table feed per rotation divided by the beam width
Pitch = (mm/rot) / (N x T)
If all other parameters are held constant: CTDIVOL µ 1 / pitch
CTDIVOL may not change in the expected manner if the scanner automatically adjust other parameters when the pitch is changed
CTDIVOL is constant when pitch is changed due to changes to other parameters for GE, Philips and Siemens
The relationship between CTDIVOL and pitch depends on scan mode or software version for Toshiba

24. Dose Reduced as Pitch Increases
Beam Width has some overlap at each view angle from rotation to rotation
Pitch = 1
No overlap of Beam Width at each view angle and no view angles not covered at certain table positions
Pitch > 1
Some view angles are not covered by the beam width at certain table positions

25. Tube Rotation Time
Is the length of time, in seconds, that the X-ray beam is “on” during a gantry rotation
It takes into account the gantry rotation time and angular acquisition range
If all other parameters are held constant:
CTDIVOL µ Exposure Time per Rotation
The scanner may or may not automatically compensate for changes in Exposure Time per Rotation(for example, by changing the tube current)
CTDIVOL is independent of Exposure Time per Rotation: GE, Siemens, Philips, Toshiba (AEC)

26. Tube Current (mA) Effective mAs = mAs / pitch
Determines the number of electrons accelerated across the x-ray tube per unit time; Units: milliAmperes (mA)
Tube Current Time Product (mAs) is the product of Tube Current and the Exposure Time per Rotation
Effective mAs = mAs / pitch
CTDIVOL is directly proportional to mA, mAs and/or effective mAs

27. Tube Potential (kVp)
Is the electrical potential applied across the x-ray tube to accelerate electrons toward the target material
Units: kiloVolts (kV or kVp)
CTDIVOL is approximately proportional to the square of the percentage change in Tube Potential
n ≈ 2 to 4

28. Beam Shaping (Bow Tie) Filter
Field of View (FOV)
Is the diameter of the primary beam in the axial plane at the gantry isocenter; units: millimeters (mm)
CTDIVOL may decrease with a decrease in the FOV
The relationship is vendor specific
Beam Shaping (Bow Tie) Filter
modifies the energy spectrum & spatial distribution of the primary beam (may include a bow tie filter and/or flat filters)
CTDIVOL is affected by a change in Beam Shaping Filters
The relationship is vendor and filter specific
**Users should monitor the CTDIVOL values when changing the Field of View or Beam Shaping Filters

29. Dose Reduction Techniques
The acquisition section thickness µ Signal to Noise Ratio (SNR)
As beam width ↑ SNR increases ↑
This allows for dose reduction in that the same SNR can be achieved by using a lower mAs (lower dose)
And vice versa: reducing the acquisition beam width reduces the dose efficiency, thus increasing CTDIVOL with more significant increases in dose seen for the Body (32 cm) phantom than in the Head (16 cm) phantom
∆ beam width
Head
Body
20 → 10 mm
~ 15%
~ 20%
10 → 5 mm
~ 35%
~ 38%
20 → 5 mm
~ 55%
~ 65%

30. Dose Reduction Techniques
kVp: Dose Reduction with regard to clinical applications
Lower kVp improves vascular enhancement (Iodine k-edge)
Keep in mind lower kVp results in a less penetrable beam which could lead to artifacts in larger patients (Trade off: Dose vs. Image Quality)
mAs selection with regard to clinical application
Lung: high contrast does NOT require low noise
Liver: need for low contrast = need for low noise
50% reduction mAs (dose) = 41% increase in noise
∆ kVp
dose reduction
140 → 120
~ 32%
120 → 80
~ 64%

31. Dose Modulation and Reduction
Many CT scanners automatically adjust the technique parameters (and as a result the CTDIVOL) to achieve a desired level of image quality and/or to reduce dose
Thinner patient sections: reduce mA thus reducing dose
Thicker patient sections: increase mAs (dose) thus reducing photon starvation artifacts (i.e. improved image quality)
Dose Modulation and Reduction techniques vary by scanner manufacturer, model and software version
GE auto mA & smart mA
Siemens CARE kV & CARE Dose
Toshiba Sure Exposure
Philips Dose Right

32. Automatic Exposure Control (AEC)
Automatically adapts the Tube Current (mA) or Tube Potential (kVp) according to patient attenuation to achieve a specified image quality
Centering the patient in the gantry is VITAL for most AEC systems
AEC aims to deliver a specified image quality across a range of patient sizes. It tends to increase CTDIVOL for large patients and decrease it for small patients relative to a reference image quality parameter
The use of AEC may decrease OR increase CTDIVOL depending on the patient size and body area imaged and image quality requested
* Potential for Patient Dose Reduction through Protocol Optimization * which requires an understanding of how the manufacturer’s AEC system (tube current modulation) functions

33. Image Quality Reference Parameter
Is the AEC parameter that is set by the user to define the desired level of image quality
Changing the Image Quality Reference Parameter will affect the CTDIVOL
Increased image quality (eg. reduced noise) will increase the CTDIVOL
Allowing more noise (reduced image quality) will decrease the CTDIVOL
The effect on CTDIVOL when changing the Image Quality Reference Parameter is vendor dependent
GE Noise Index
Siemens Reference mA(s)
Toshiba Standard Deviation
Philips Reference Image

34. Rotational / Angular mA Modulation
AEC feature that adjusts the Tube Current (mA) as the x-ray tube rotates around the patient to compensate for attenuation changes with view angle
Attempt to deliver similar dose to the detector at all view angles (i.e. maintains photon flux to the detector)
Maintains image quality and noise in asymetric regions (eg. reduce photon starvation artifacts in lateral shoulder)

35. Angular Tube Current Modulation
Angular mA Modulation typically uses information from one or two view localizers
Rotational AEC w/out scout or topogram:
Patient attenuation data sent to generator control for mA modulation with a delay of 180°
“On the Fly” modulation based on attenuation data of previous rotation
Dose Modulation and Reduction

36. Longitudinal (z-axis) mA Modulation
AEC feature that adjusts the Tube Current as patient attenuation changes in the longitudinal direction
One or two CT Localizer Radiographs used to estimate patient attenuation

37. Angular & Longitudinal mA Modulation
Incorporates the properties of both Angular and Longitudinal Tube Current Modulation to:
Adjust the Tube Current based on the patient’s overall attenuation
Modulate the Tube Current in the angular (x-y) and longitudinal (z) dimensions to adapt to the patient’s shape

38. Angular & Longitudinal mA Modulation

39. Siemens CARE Dose 4D Reference mAs based AEC Optimization
Principle: different size patients require different noise level to maintain image quality
User selects a Reference Effective mAs for an Average Patient
Design defaults: Adult kg, Pediatric 20 kg (5 year old)
Optimization
User can set up various reference techniques (eg. thin, avg., obese, etc.) to minimize large scale compensation variations or reaching Min/Max
Topogram(s): provide Longitudinal (z-axis) AND Rotational (x-y) Automatic Exposure Control (i.e. 3-Dimensional)
4D “real-time” modulation uses feedback from previous rotation

40. Automatic Tube Potential Selection
AEC feature that selects the kVp according to the diagnostic task and patient size in order to achieve the desired image quality at a lower CTDIVOL
kVp is not modulated in the same fashion as Tube Current
It does not change with different tube positions (view angles) around the patient
The Tube Potential for a specific patient, anatomic region and diagnostic tasks is selected and held constant for that acquisition, though it may be changed to a different tube potential for a different diagnostic task

41. Iterative Reconstruction
This reconstruction technique uses the information acquired during the scan and repeated reconstruction steps to produce an image with less “noise” or better image quality (e.g., higher spatial resolution or decreased artifacts) than is achievable using standard reconstruction techniques
The use of Iterative Reconstruction by itself may not decrease CTDIVOL; with use of Iterative Reconstruction, image quality will change and this may allow a reduction in the CTDIVOL by adjusting the acquisition parameters used for the exam

42. Iterative Reconstruction
Iterative Reconstruction may be completed using data in Image Space, Sinogram Space or a Model Based Approach
Changing the % Level of the iterative reconstruction used may or may not affect the CTDIVOL of the scan and will affect the image quality of the final set of images
In consultation, the Radiologists and Medical Physicists at an institution may adjust the acquisition parameters for studies reconstructed using iterative reconstruction or other post processing software based on the imaging task, the patient population, the desired image quality, dose concerns and the needs of the interpreting Radiologist

43. Dose Display
Information about the CTDIVOL planned for each scan based on the selected technique parameters is typically displayed BEFORE the exam on the user console
It is important to check CTDIVOL before a study is performed to ensure that the output of the scanner is appropriate for the specific patient and diagnostic task
Information about the CTDIVOL delivered by each scan is typically reported in a data page or DICOM structured dose report AFTER the exam
It is useful to check CTDIVOL after a study is performed to ensure that the output of the scanner was as expected
Dose information provided typically also includes the DLP and the CTDI (reference) phantom size.
*** CTDIVOL is NOT a Patient Dose ***
underestimates dose for patient sizes < reference phantom
overestimates dose for patient sizes > reference phantom

44. Summing Dose Report Values
CTDIVOL values for separate series are NOT to be summed to give a “total” CTDIVOL for a study
This is especially true if the series covers different anatomical regions
DLP is typically summed over all series in the Post Study Data Page to provide an estimate of the total patient exposure
Extreme care should be taken when considering summed DLPs because different phantoms may have been used to calculate the CTDIVOL values used to determine DLP
A medical physicist should be contacted if patient specific dose estimates are required

45. Dose Notification Levels
Notification Levels may be set on a CT scanner for each series within an exam protocol
If the planned CTDIVOL is above the Notification Level and triggers the notification, the user has the opportunity to edit or confirm the technique settings
Notification Levels may be exceeded when appropriate for a specific patient or diagnostic task (e.g., in very large patients or contrast bolus monitoring scans)

46. Dose Alert Levels
Dose Alert Levels require specific action by the operator to continue scanning
Dose Alert Levels are typically much higher than Notification Levels and take into account all series within the exam
Triggering a Dose Alert requires that the operator confirm the protocol and settings are correct by entering in his or her name. Optionally, sites may require that the operator provide a brief explanation in the provided field

47. Radiation Dose Structured Reports
Radiation Dose Structured Reports (RDSRs) are provided in newer software versions in a defined DICOM format
They provide the most complete set of information regarding the irradiating events
The reports are very detailed and require an RDSR viewer for easy visualization of relevant information

48. References Common CT Protocols for specific for each vendor
Common CT Protocols for specific for each vendor
General radiation safety and CT Dose Education slides including education material specific for each CT manufacturer
Dose Alert Reference Levels

49. References http://www.radiologyinfo.org/ http://www.acr.org/

50. Instructions to the Technologist
Increase awareness for the need to decrease radiation dose to children during CT scanning. Encourage your fellow professionals to get involved in the effort.
Be committed to make a change in your daily practice by working as a team with your radiologist, physicist, referring doctors and parents to decrease the radiation dose. Sign the pledge! at
Know your ASRT Computed Tomography Practice Standards. Standards 1 and 2 on assessment and analysis are your guide to ensuring an appropriate action plan is established for completing a CT exam.
Work with your physicist, radiologist and department manager to review your adult CT protocols; adult-size kV and mAs are not necessary for small bodies. Pediatric protocols need be “down-sized” for children.
Be involved with your patients. Be the patient’s advocate. Ask the questions required to ensure that you “child-size” the scan and only scan the area required to obtain the necessary information

51. Pediatric CT Considerations
Need for Higher Contrast to Noise Ratio (CNR)
Less adipose tissue than adults (naturally less subject contrast)
Smaller Volume: can use thinner slices
Use Faster Rotation Speed
Limit patient motion artifacts
Children more radiosensitive than adults; longer time to experience induced effects of ionizing radiation exposure
Need for reduction in dose; however as mAs ↓ Noise ↑
Reduce kVp & increase mAs (improves image quality) if CTDIVOL is maintained

52. Estimated Organ Doses and Lifetime Cancer Risk Increased for Pediatric Patients undergoing CT exams
Children are more radiosensitive than adults and exposures at younger ages contribute to an increase in lifetime cancer risk

53. “Child Sizing” CT Dose Need for Size Specific Pediatric protocols
Weight based protocols are better than aged based protocols especially for Chest, Abdomen and Pelvic CT examinations
Weight + Height or Patient Diameter based protocols would be ideal
Age based protocols for Head exams are acceptable
ACR Reference Levels / Dose Limit
Pediatric Head (1 year): 35 / 40 mGy
Ped Abdomen (5 year): 15 / 20 mGy
Image Gently tables suggest scaling adult protocols in order to maintain dose; this is a minimum dose saving method; further optimization is recommended

56. Displayed CTDIVOL in Pediatric CT
*** CTDIVOL is NOT a Patient Dose ***
underestimates dose for patient sizes < reference phantom
overestimates dose for patient sizes > reference phantom
CTDIVOL can often underestimate the Pediatric Abdomen dose by 50% or greater
Most manufacturers use the 32 cm diameter phantom as a reference for Pediatric Body CT scans

57. Effective Dose (mSv) DLP (mGy·cm) = CTDIVOL (mGy) x Scan Length (cm)
Dose Length Product (DLP): absorbed dose integrated over the scan length
DLP (mGy·cm) = CTDIVOL (mGy) x Scan Length (cm)
reflects the total energy absorbed (potential biological effect)
Effective Dose (E): reflects the risk of a non-uniform exposure in terms of an equivalent whole body exposure
E (mSv) ≈ DLP x k-factor
where k-factor (mSv/mGy·cm) is based on the Radiosensitivity of the organs in the scan volume
k (Abd/Pel) > k (Head/Neck)
k (Pediatric) > k (Adult)

58. Common Misuse of Effective Dose
Much like CTDIVOL, Effective Dose is NOT a patient dose
E reflects the risk of a non-uniform exposure in terms of an equivalent whole body exposure and was defined for quantifying the risk to Adult Radiation Workers
It was NOT designed to represent Children
It is NOT a metric for quantifying Cancer Risk Estimates for individual Diagnostic Exams.

59. Effective Dose
Effective dose is best used to optimize exams and to compare risks between proposed exams

60. Benefits of CT Radiation Risk vs. Medical Benefit
Don’t let the benefit of quality diagnostic imaging get lost in the general fear of radiation exposure
“If an institution performs 300 CT scans per year, the risk benefit equation balances if CT saves 1 life every 4 years”
Haaga AJR 2001; 177:
What to tell the patient?
Reassure them that the exam is necessary for proper treatment and diagnosis

61. Conclusions Radiation Risk vs. Medical Benefit
As Low As Diagnostically Acceptable
There is no dose limit for patients undergoing CT examinations in the U.S.; therefore, it is the responsibility of physicians, physicists, and technologists to limit the dose to the patient
Techniques should be minimized without deducting from image quality
Reduce unnecessary exams & series
Optimize protocols in terms of Patient Dose & Image Quality
Image Gently
Development of Patient size-specific protocols & techniques

62. Questions
Please contact the medical physicist providing support for your CT practice, your lead technologist, supervising radiologist or manufacturer’s application specialist with questions regarding these important topics and concepts.