Computed Tomography Dosimetry & Protocol Optimization

ABMP Medical Health Physics Exam

Track your progress as you read

Finished sections get a check mark.

1 Radiation Detection Instrumentation

Ionization Chambers & Geiger-Mueller Counters Scintillation Detectors, Photomultipliers & Gamma Spectroscopy Liquid Scintillation & Gamma Counting Systems: Well Counters & Thyroid Probes Personnel Dosimeters & Electronic Radiation Monitors Neutron Detection & Emergency Field Instrumentation

2 Laboratory Design & Contamination Control

Radiopharmaceutical Laboratory Design & USP <825> Standards Ventilation Systems, Fume Hoods & Contamination Engineering

3 Radiation Protection Standards & Dose Limits

Federal Regulatory Framework & Advisory Authorities Federal Guidance Documents: NRC Regulatory Guides, NUREGs, EPA & FDA Guidance Occupational & Public Dose Limits and ALARA Principles Radiobiological Basis of Limits & Risk Communication

4 Medical Licensing, Written Directives & Security

NRC Form 313 Authorizations, RSO Responsibilities & Written Directives Medical Event Reporting & 10 CFR 37 Category 1/2 Source Security

5 Diagnostic Radiology Structural Shielding

Diagnostic Shielding Methodology: NCRP Report 147 Modality Shielding: Radiography, Fluoroscopy & CT Facilities

6 Megavoltage Therapy Vault & Facility Shielding

Megavoltage Vault Primary & Secondary Barriers: NCRP Report 151 Therapy Maze Design, Photoneutrons & Shielded Doors PET/CT & Nuclear Medicine Facility Shielding (AAPM TG-108)

7 Radiation Oncology Quality Management & Safety

External Beam QA & Dosimetry Calibration Protocols Brachytherapy Safety, Inventory Control & Leak Testing

8 Radiation Oncology Emergency Procedures & TPS QA

Treatment Planning System (TPS) QA & Patient Verification Radiation Oncology Emergency Procedures: HDR Stuck Sources & Linac Failures Medical Events, Incident Investigation & Joint Commission Requirements in Radiation Oncology

9 Diagnostic Imaging Safety & Fluoroscopy Management

Diagnostic X-ray & Fluoroscopy Safety Interventional Fluoroscopy Dose Management & NCRP Report 168

10 CT Optimization, MQSA & Diagnostic Accreditation

Computed Tomography Dosimetry & Protocol Optimization Mammography Quality Standards Act (MQSA) & Mammography Physics Diagnostic Imaging Quality Management, Accreditation & Regulatory Oversight

11 Nuclear Medicine Radiopharmacy & Unsealed Therapy

Radiopharmaceutical QC & Radionuclide Calibrators Unsealed Source Therapy: I-131 Protocols & Patient Release Criteria

12 Targeted Radionuclide Therapy, Decontamination & RAM Transport

Targeted Radionuclide Therapies: Lu-177, Ra-223, Y-90 & Radioimmunotherapy Contamination Surveys, Spill Response & Personnel Decontamination Radioactive Waste Disposal & US DOT Transport Regulations

13 Radiobiology & Radiation Risk Assessment

Cellular Radiobiology, DNA Damage & Dose-Response Models Prenatal Radiation Risks & Population Exposures

14 Non-Ionizing Radiation Safety: Lasers & MRI

Healthcare Laser Safety: ANSI Z136.1 & Z136.3 MRI Safety, Facility Design & Bioeffects

Computed Tomography Dosimetry & Protocol Optimization

Key Takeaways Explain This Topic AI

CTDI100 integrates the radiation dose profile along a 100 mm pencil ionization chamber inside 16 cm (head) or 32 cm (body) PMMA phantoms; weighted CTDI (CTDIw) computes the radial average (1/3 center + 2/3 periphery), and volume CTDI (CTDIvol = CTDIw / pitch) accounts for helical overlap or gaps.

Dose-Length Product (DLP = CTDIvol × L) represents the total radiation energy imparted along the irradiated anatomical length, enabling effective dose estimation (E = k × DLP) using normalized ICRP anatomical conversion coefficients.

Size-Specific Dose Estimate (SSDE = f_size × CTDIvol, AAPM Reports 204 and 220) corrects for patient size using water-equivalent diameter; relative to the 32 cm phantom, f_size is about 2.06 at 16 cm, 1.14 at 32 cm, and 0.79 at 42 cm.

Protocol optimization relies on automatic tube current modulation (longitudinal z-axis and angular x-y modulation), reducing tube potential (70–100 kVp) to optimize iodine contrast near its 33.2 keV k-edge, and deploying iterative (IR) or deep learning reconstruction (DLR) to lower dose while preserving spatial resolution and noise texture.

The ACR CT phantom (Gammex 464) checks CT number accuracy (for example water −7 to +7 HU and air −1005 to −970 HU), low-contrast detectability by contrast-to-noise ratio, uniformity (edges within ±5 HU of center), and spatial resolution, alongside CTDIvol reference and pass/fail values.

Last updated: October 2026

Computed Tomography Dosimetry Physics & Phantoms

Computed Tomography (CT) delivers higher organ-specific radiation doses than conventional projection radiography, accounting for the majority of medical ionizing radiation exposure in developed nations. Rigorous CT dosimetry requires standardized metrics that reflect scanner radiation output under reproducible geometric conditions. The foundation of modern CT dosimetry rests on the pencil-shaped ionization chamber and cylindrical polymethyl methacrylate (PMMA) phantoms.

The 100 mm Pencil Ionization Chamber & Dosimetric Integration

Conventional small-volume ion chambers fail to capture the extensive scattered radiation tails characteristic of narrow or wide x-ray fans interacting within scattering media. CT dosimetry therefore utilizes a specialized pencil ionization chamber with an active length of exactly 100 mm 100\text{ mm} 100 mm ( 10 cm 10\text{ cm} 10 cm ) and an active volume of approximately 3 cm 3 3\text{ cm}^3 3 cm 3 (outer diameter ≈ 9 mm \approx 9\text{ mm} ≈ 9 mm ).

The pencil chamber measures the integral of the dose profile D ( z ) D(z) D ( z ) along the longitudinal rotation axis ( z z z -axis). The fundamental dosimetric quantity is the Computed Tomography Dose Index 100 ( C T D I 100 CTDI_{100} CT D I 100 ​ ) , representing the accumulated dose over a 100 mm 100\text{ mm} 100 mm integration length for a single axial gantry rotation, normalized by the total nominal beam collimation width:

C T D I 100 = 1 N ⋅ T ∫ − 50 mm + 50 mm D ( z ) d z CTDI_{100} = \frac{1}{N \cdot T} \int_{-50\text{ mm}}^{+50\text{ mm}} D(z) \, dz CT D I 100 ​ = N ⋅ T 1 ​ ∫ − 50 mm + 50 mm ​ D ( z ) d z

where:

N N N is the number of active data channels (detector rows simultaneously collecting projections).

T T T is the nominal tomographic section thickness per active channel.

The product N ⋅ T N \cdot T N ⋅ T is the total nominal collimated beam width (in mm).

In practice, using an electrometer reading M M M , calibration factor N k N_k N k ​ (air kerma calibration factor), temperature-pressure correction factor C T P C_{TP} C TP ​ , and pencil chamber active length L = 100 mm L = 100\text{ mm} L = 100 mm :

C T D I 100 = M ⋅ N k ⋅ C T P ⋅ L N ⋅ T CTDI_{100} = \frac{M \cdot N_k \cdot C_{TP} \cdot L}{N \cdot T} CT D I 100 ​ = N ⋅ T M ⋅ N k ​ ⋅ C TP ​ ⋅ L ​

PMMA Cylindrical Dosimetry Phantoms

To measure C T D I 100 CTDI_{100} CT D I 100 ​ under realistic scattering conditions, measurements are performed in standard cylindrical polymethyl methacrylate (PMMA, acrylic, density ρ = 1.19 g/cm 3 \rho = 1.19\text{ g/cm}^3 ρ = 1.19 g/cm 3 ) phantoms. Two standardized cylinder sizes are codified under IEC 60601-2-44 and FDA 21 CFR 1020.33:

Head / Pediatric Body Phantom : 16 cm 16\text{ cm} 16 cm diameter, 14 – 15 cm 14\text{--}15\text{ cm} 14 – 15 cm length, representing the adult head and pediatric body/torso.

Adult Body Phantom : 32 cm 32\text{ cm} 32 cm diameter, 14 – 15 cm 14\text{--}15\text{ cm} 14 – 15 cm length, representing the adult abdomen, pelvis, and thorax.

Each phantom contains five parallel cylindrical bore holes: one central hole along the rotational cylinder axis ( z z z -axis) and four peripheral holes located 10 mm 10\text{ mm} 10 mm beneath the outer cylinder surface at 12, 3, 6, and 9 o'clock. When a bore is not housing the pencil chamber, it must be filled with a solid PMMA acrylic plug to preserve scattering geometry.

CTDI Metrics: From C T D I 100 CTDI_{100} CT D I 100 ​ to C T D I w CTDI_w CT D I w ​ , C T D I vol CTDI_{\text{vol}} CT D I vol ​ , and DLP

Weighted CTDI ( C T D I w CTDI_w CT D I w ​ )

Radiation dose within a cylindrical phantom is non-uniform due to beam attenuation and scatter buildup. In the 32 cm 32\text{ cm} 32 cm body phantom, the peripheral dose is typically two to three times greater than the central dose ( C T D I 100 , periphery / C T D I 100 , center ≈ 2 – 3 CTDI_{100,\text{periphery}} / CTDI_{100,\text{center}} \approx 2\text{--}3 CT D I 100 , periphery ​ / CT D I 100 , center ​ ≈ 2 – 3 ). In the 16 cm 16\text{ cm} 16 cm head phantom, attenuation is less severe, yielding a more uniform profile ( C T D I 100 , periphery / C T D I 100 , center ≈ 1.1 – 1.2 CTDI_{100,\text{periphery}} / CTDI_{100,\text{center}} \approx 1.1\text{--}1.2 CT D I 100 , periphery ​ / CT D I 100 , center ​ ≈ 1.1 – 1.2 ).

To establish an average dose within the scan plane, the Weighted CTDI ( C T D I w CTDI_w CT D I w ​ ) applies a linear spatial weighting approximation, assigning 1 / 3 1/3 1/3 weight to the central core and 2 / 3 2/3 2/3 weight to the peripheral annulus:

C T D I w = 1 3 C T D I 100 , center + 2 3 C T D I 100 , periphery CTDI_w = \frac{1}{3} CTDI_{100,\text{center}} + \frac{2}{3} CTDI_{100,\text{periphery}} CT D I w ​ = 3 1 ​ CT D I 100 , center ​ + 3 2 ​ CT D I 100 , periphery ​

where C T D I 100 , periphery CTDI_{100,\text{periphery}} CT D I 100 , periphery ​ is the arithmetic mean of the four peripheral measurements (or an average of the 12 o'clock and 6 o'clock positions in symmetric bow-tie configurations):

C T D I 100 , periphery = C T D I 12 + C T D I 3 + C T D I 6 + C T D I 9 4 CTDI_{100,\text{periphery}} = \frac{CTDI_{12} + CTDI_3 + CTDI_6 + CTDI_9}{4} CT D I 100 , periphery ​ = 4 CT D I 12 ​ + CT D I 3 ​ + CT D I 6 ​ + CT D I 9 ​ ​

Volume CTDI ( C T D I vol CTDI_{\text{vol}} CT D I vol ​ ) & Pitch Dynamics

In helical (spiral) CT, the patient couch translates continuously while the gantry rotates. The relationship between couch translation per rotation and collimated beam width is defined as the pitch :

pitch = Δ d N ⋅ T \text{pitch} = \frac{\Delta d}{N \cdot T} pitch = N ⋅ T Δ d ​

where Δ d \Delta d Δ d is couch translation distance per 360 ∘ 360^\circ 36 0 ∘ gantry rotation. The Volume CTDI ( C T D I vol CTDI_{\text{vol}} CT D I vol ​ ) accounts for helical pitch by scaling C T D I w CTDI_w CT D I w ​ inversely with pitch:

C T D I vol = C T D I w pitch = C T D I w × ( N ⋅ T Δ d ) CTDI_{\text{vol}} = \frac{CTDI_w}{\text{pitch}} = CTDI_w \times \left( \frac{N \cdot T}{\Delta d} \right) CT D I vol ​ = pitch CT D I w ​ ​ = CT D I w ​ × ( Δ d N ⋅ T ​ )

Pitch < 1.0 < 1.0 < 1.0 (Overlapping Helices) : Radiation beams overlap across contiguous rotations, increasing dose ( C T D I vol > C T D I w CTDI_{\text{vol}} > CTDI_w CT D I vol ​ > CT D I w ​ ).

Pitch = 1.0 = 1.0 = 1.0 (Contiguous Scanning) : C T D I vol = C T D I w CTDI_{\text{vol}} = CTDI_w CT D I vol ​ = CT D I w ​ .

Pitch > 1.0 > 1.0 > 1.0 (Extended Pitch / Gaps) : X-ray beams are spaced out, decreasing dose ( C T D I vol < C T D I w CTDI_{\text{vol}} < CTDI_w CT D I vol ​ < CT D I w ​ ).

Important

C T D I vol CTDI_{\text{vol}} CT D I vol ​ is NOT patient dose . C T D I vol CTDI_{\text{vol}} CT D I vol ​ is a standardized measure of scanner radiation output inside a rigid, homogeneous 16 cm 16\text{ cm} 16 cm or 32 cm 32\text{ cm} 32 cm PMMA cylinder. It provides a standardized metric for quality control, regulatory benchmarking, and scanner comparison, but does not represent absorbed dose to an individual patient.

Dose-Length Product ( D L P DLP D L P ) & Effective Dose ( E E E )

The Dose-Length Product (DLP) integrates C T D I vol CTDI_{\text{vol}} CT D I vol ​ over the entire irradiated anatomical length of the examination:

D L P = C T D I vol × L DLP = CTDI_{\text{vol}} \times L D L P = CT D I vol ​ × L

where L L L is the irradiated scan length along the z z z -axis (in cm \text{cm} cm ), yielding units of mGy ⋅ cm \text{mGy}\cdot\text{cm} mGy ⋅ cm . Note that irradiated scan length L L L exceeds the reconstructed image length due to helical overranging (the extra rotation required before and after the planned volume to interpolate images at the edges of the reconstruction volume).

Effective dose ( E E E ) (in mSv \text{mSv} mSv ) estimates broad population stochastic risk, calculated from DLP using normalized anatomy-specific conversion coefficients ( k k k ):

E = k × D L P E = k \times DLP E = k × D L P

Commonly used adult k k k -factors (in mSv / ( mGy ⋅ cm ) \text{mSv}/(\text{mGy}\cdot\text{cm}) mSv / ( mGy ⋅ cm ) ) from AAPM Report 96, which were derived with ICRP 60 tissue weighting factors, include:

• Head : k ≈ 0.0021 k \approx 0.0021 k ≈ 0.0021
• Head & neck : k ≈ 0.0031 k \approx 0.0031 k ≈ 0.0031
• Neck : k ≈ 0.0059 k \approx 0.0059 k ≈ 0.0059
• Chest : k ≈ 0.014 k \approx 0.014 k ≈ 0.014
• Abdomen & Pelvis : k ≈ 0.015 k \approx 0.015 k ≈ 0.015

Pediatric k k k -coefficients are substantially higher (e.g., pediatric chest ≈ 0.026 – 0.039 \approx 0.026\text{--}0.039 ≈ 0.026 – 0.039 , pediatric abdomen ≈ 0.020 – 0.049 \approx 0.020\text{--}0.049 ≈ 0.020 – 0.049 ) due to smaller body mass and elevated tissue radiosensitivity.

• Size-Specific Dose Estimate (SSDE): AAPM Reports 204 & 220
• Discrepancy Between Patient Habitus and Fixed Phantoms

Because C T D I vol CTDI_{\text{vol}} CT D I vol ​ is calibrated to fixed 16 cm 16\text{ cm} 16 cm or 32 cm 32\text{ cm} 32 cm cylinders, it exhibits severe physical divergence when applied to actual patients of differing sizes:

In small pediatric patients, less tissue attenuation occurs, meaning absorbed organ dose is substantially higher than the reported C T D I vol CTDI_{\text{vol}} CT D I vol ​ (often 2 × to 3 × 2\times\text{ to }3\times 2 × to 3 × higher when based on the 32 cm 32\text{ cm} 32 cm phantom).

In bariatric patients, self-shielding by peripheral adipose tissue means internal organ dose is lower than the reported C T D I vol CTDI_{\text{vol}} CT D I vol ​ .

AAPM Report 204: Geometric Dimension Conversion

To bridge this gap, AAPM Task Group 204 formulated the Size-Specific Dose Estimate (SSDE) , scaling C T D I vol CTDI_{\text{vol}} CT D I vol ​ by an exponential size conversion factor ( f size f_{\text{size}} f size ​ ):

S S D E = f size × C T D I vol SSDE = f_{\text{size}} \times CTDI_{\text{vol}} SS D E = f size ​ × CT D I vol ​

Report 204 derives f size f_{\text{size}} f size ​ from physical patient dimensions measured on axial CT images or scout radiographs (topograms):

• Anterior-Posterior (AP) dimension
• Lateral (LAT) dimension

Sum of dimensions ( A P + L A T AP + LAT A P + L A T ) or geometric mean diameter ( D geom = A P × L A T D_{\text{geom}} = \sqrt{AP \times LAT} D geom ​ = A P × L A T ​ )

AAPM Report 220: Water-Equivalent Diameter ( D w D_w D w ​ )

Geometric dimensions work poorly in the thorax because lung contains air (CT numbers around − 700 to − 900 HU -700\text{ to }-900\text{ HU} − 700 to − 900 HU ) and attenuates far less than the same thickness of water. AAPM Report 220 therefore recommends the attenuation-based Water-Equivalent Diameter ( D w D_w D w ​ ) calculated directly from CT slice pixel values:

A w = [ CT ‾ R O I 1000 + 1 ] A R O I A_w = \left[ \frac{\overline{\text{CT}}_{ROI}}{1000} + 1 \right] A_{ROI} A w ​ = [ 1000 CT RO I ​ ​ + 1 ] A RO I ​

D w = 2 A w π = 2 [ CT ‾ R O I 1000 + 1 ] A R O I π D_w = 2 \sqrt{\frac{A_w}{\pi}} = 2 \sqrt{ \left[ \frac{\overline{\text{CT}}_{ROI}}{1000} + 1 \right] \frac{A_{ROI}}{\pi} } D w ​ = 2 π A w ​ ​ ​ = 2 [ 1000 CT RO I ​ ​ + 1 ] π A RO I ​ ​ ​

where CT ‾ R O I \overline{\text{CT}}_{ROI} CT RO I ​ is the mean CT number (in Hounsfield Units) within the patient contour, and A R O I A_{ROI} A RO I ​ is the cross-sectional area of the patient contour. The size conversion factor is parameterized as:

f size = a ⋅ e − b ⋅ D w f_{\text{size}} = a \cdot e^{-b \cdot D_w} f size ​ = a ⋅ e − b ⋅ D w ​

For the 32 cm 32\text{ cm} 32 cm reference phantom ( a = 3.704369 a = 3.704369 a = 3.704369 , b = 0.03671937 cm − 1 b = 0.03671937\text{ cm}^{-1} b = 0.03671937 cm − 1 ):

D w = 16 cm D_w = 16\text{ cm} D w ​ = 16 cm (small child): f size ≈ 2.06 f_{\text{size}} \approx 2.06 f size ​ ≈ 2.06 , so the patient dose is about twice the reported C T D I vol CTDI_{\text{vol}} CT D I vol ​ .

D w = 32 cm D_w = 32\text{ cm} D w ​ = 32 cm : f size ≈ 1.14 f_{\text{size}} \approx 1.14 f size ​ ≈ 1.14 , not 1.00 1.00 1.00 , because a water-equivalent patient attenuates less than the PMMA phantom; f size f_{\text{size}} f size ​ reaches 1.0 1.0 1.0 near 36 cm 36\text{ cm} 36 cm .

D w = 42 cm D_w = 42\text{ cm} D w ​ = 42 cm (large adult): f size ≈ 0.79 f_{\text{size}} \approx 0.79 f size ​ ≈ 0.79 , so the patient dose is lower than the reported C T D I vol CTDI_{\text{vol}} CT D I vol ​ .

• CT Protocol Optimization: ATCM, kVp Selection & Advanced Reconstruction
• Automatic Tube Current Modulation (ATCM)

ATCM continuously modulates tube current (mA) to maintain uniform image quality across varying patient cross-sections, avoiding excessive dose in thin regions and photon starvation in dense regions:

Longitudinal Modulation ( z z z -axis) : Adjusts tube current along the patient's long axis based on local attenuation profiles derived from the scout/topogram. Tube current decreases through the low-attenuation thorax and increases through the dense shoulder girdle and bony pelvis.

Angular Modulation ( x - y x\text{-}y x - y axis) : Modulates tube current within each 360 ∘ 360^\circ 36 0 ∘ rotation. Because lateral attenuation through the torso is 2 × to 4 × 2\times\text{ to }4\times 2 × to 4 × higher than anterior-posterior (AP) attenuation, ATCM peaks tube current during lateral projections and reduces it during AP/PA projections.

Vendor Implementations : Manufacturers employ distinct target image quality metrics: Noise Index (GE), Quality Reference mAs (Siemens), Standard Deviation (Canon/Toshiba), or Dose Right Index (Philips). Physicists must understand that increasing Noise Index on GE systems decreases radiation dose, whereas increasing Quality Reference mAs on Siemens systems increases radiation dose.

Tube Potential (kVp) Selection & Iodine Contrast Optimization

The photoelectric effect probability scales as Z 3 / E 3 Z^3 / E^3 Z 3 / E 3 , where Z Z Z is atomic number and E E E is photon energy. The k-shell absorption edge of iodine is 33.2 keV 33.2\text{ keV} 33.2 keV .

Standard adult scanning traditionally utilized 120 kVp 120\text{ kVp} 120 kVp , producing a mean beam energy of ≈ 60 – 70 keV \approx 60\text{--}70\text{ keV} ≈ 60 – 70 keV , well above the iodine k-edge.

Lowering tube potential to 70 – 100 kVp 70\text{--}100\text{ kVp} 70 – 100 kVp shifts the polyenergetic spectrum downward toward 33.2 keV 33.2\text{ keV} 33.2 keV , dramatically elevating the mass attenuation coefficient of iodinated contrast media.

In CT angiography (CTA) and contrast-enhanced abdominal imaging, reducing kVp enhances vascular and parenchymal iodine contrast, enabling a 30 – 50 % 30\text{--}50\% 30 – 50% reduction in radiation dose or a reduction in total administered contrast volume.

Pitfall : In large adult patients, lower kVp causes severe photon starvation and beam hardening streaks; tube potential must be matched to patient habitus ( 70 – 80 kVp 70\text{--}80\text{ kVp} 70 – 80 kVp for infants/small children, 80 – 100 kVp 80\text{--}100\text{ kVp} 80 – 100 kVp for normal-weight adults, 120 – 140 kVp 120\text{--}140\text{ kVp} 120 – 140 kVp for bariatric patients).

Iterative Reconstruction (IR) vs. Filtered Back Projection (FBP)

Filtered Back Projection (FBP) : Fast, non-iterative analytic algorithm using a ramp filter to de-blur 1 / r 1/r 1/ r backprojections. FBP assumes noiseless projection data and point-like focal spots. When radiation dose (photon flux) is reduced, FBP exhibits severe streak artifacts and elevated image noise.

Statistical Iterative Reconstruction (SIR) : Models photon counting statistics (Poisson distribution) and electronic detector noise (Gaussian distribution). SIR iteratively refines estimated cross-sectional slices by comparing simulated forward projections against measured raw data, achieving a 20 – 40 % 20\text{--}40\% 20 – 40% dose reduction at matched noise.

Model-Based Iterative Reconstruction (MBIR) : Expands modeling to full system optics (finite focal spot dimensions, beam geometry, detector element response, and three-dimensional scatter). MBIR permits 50 – 75 % 50\text{--}75\% 50 – 75% dose reductions but is computationally intensive and can alter the noise power spectrum (NPS), producing an unfamiliar "plastic" or "waxy" image texture.

Deep Learning Reconstruction (DLR) : Employs deep convolutional neural networks trained on high-dose, ultra-high-resolution FBP and MBIR dataset pairs. DLR eliminates low-dose streak artifacts and restores a natural, fine-grained noise texture matching high-dose FBP while achieving up to 60 – 70 % 60\text{--}70\% 60 – 70% dose reduction without spatial resolution loss.

Image Gently & Image Wisely Campaigns

Image Gently : Run by the Alliance for Radiation Safety in Pediatric Imaging, Image Gently promotes child-sized CT protocols: eliminating adult kVp/mAs settings on children, using pediatric-specific bow-tie filters, activating ATCM with pediatric reference parameters, and restricting multiphase acquisitions.

Image Wisely : Joint initiative of the ACR, RSNA, AAPM, and ASRT to reduce unnecessary adult examinations, optimize imaging protocols, and ensure examinations adhere to the ALARA principle.

• American College of Radiology (ACR) CT Accreditation Program
• The ACR CT Accreditation Phantom (Model 464)

The ACR CT Accreditation Program verifies that clinical scanners meet baseline quality benchmarks for image quality, spatial resolution, contrast detectability, and radiation dose. The standard testing tool is the Gammex Model 464 ACR CT Accreditation Phantom ( 20 cm 20\text{ cm} 20 cm diameter solid PMMA-based cylinder) composed of four distinct contiguous 4 cm 4\text{ cm} 4 cm modules:

Module Purpose ACR Criteria (selected)

1: CT number accuracy, positioning, slice width Inserts of known material plus wire ramps Water − 7 -7 − 7 to + 7 HU +7\text{ HU} + 7 HU ; air − 1005 -1005 − 1005 to − 970 HU -970\text{ HU} − 970 HU ; polyethylene − 107 -107 − 107 to − 84 HU -84\text{ HU} − 84 HU ; acrylic + 110 +110 + 110 to + 135 HU +135\text{ HU} + 135 HU ; bone-equivalent + 850 +850 + 850 to + 970 HU +970\text{ HU} + 970 HU

2: Low-contrast detectability Rod groups of 2–6 mm (and 25 mm) diameter at about 0.6% (6 HU) contrast Contrast-to-noise ratio above 1.0 1.0 1.0 for adult head and adult abdomen protocols, above 0.7 0.7 0.7 for pediatric head, and above 0.4 0.4 0.4 for pediatric abdomen

3: Uniformity and noise Uniform tissue-equivalent material Edge ROIs within ± 5 HU \pm 5\text{ HU} ± 5 HU of the center (5–7 HU is a minor and more than 7 HU a major deficiency)

4: High-contrast (spatial) resolution Aluminum bar patterns of 4, 5, 6, 7, 8, 9, 10, and 12 lp/cm Scored against ACR limits for each protocol

To be accredited, a unit must pass both the clinical image review and the phantom image review, and a qualified medical physicist must perform the annual system performance evaluation, oversee the technologist QC program, and perform the dosimetry.

ACR Diagnostic Reference Levels (DRL) and Pass/Fail Dose Thresholds

Facilities must submit dosimetric measurements performed using the 100 mm 100\text{ mm} 100 mm pencil chamber and standard 16 cm 16\text{ cm} 16 cm and 32 cm 32\text{ cm} 32 cm PMMA phantoms. The ACR establishes both Reference Levels (advisory benchmarks representing the 75th percentile of clinical practice) and Pass/Fail Thresholds (maximum allowable dose ceilings):

Examination Category Standard Dosimetry Phantom ACR Reference Level ( C T D I vol CTDI_{\text{vol}} CT D I vol ​ ) ACR Pass/Fail Ceiling ( C T D I vol CTDI_{\text{vol}} CT D I vol ​ )

Adult Head (Brain) 16 cm 16\text{ cm} 16 cm Head Phantom 75 mGy 75\text{ mGy} 75 mGy 80 mGy 80\text{ mGy} 80 mGy

Adult Abdomen (Pelvis) 32 cm 32\text{ cm} 32 cm Body Phantom 25 mGy 25\text{ mGy} 25 mGy 30 mGy 30\text{ mGy} 30 mGy

Pediatric Head (1-year-old) 16 cm 16\text{ cm} 16 cm Head Phantom 35 mGy 35\text{ mGy} 35 mGy 40 mGy 40\text{ mGy} 40 mGy

Pediatric Abdomen (40–50 lb, about 18–23 kg) 16 cm 16\text{ cm} 16 cm phantom (or 32 cm 32\text{ cm} 32 cm ) 15 mGy 15\text{ mGy} 15 mGy ( 16 cm 16\text{ cm} 16 cm ) / 7.5 mGy 7.5\text{ mGy} 7.5 mGy ( 32 cm 32\text{ cm} 32 cm ) 20 mGy 20\text{ mGy} 20 mGy ( 16 cm 16\text{ cm} 16 cm ) / 10 mGy 10\text{ mGy} 10 mGy ( 32 cm 32\text{ cm} 32 cm )

• Worked Numerical Examples: Clinical CT Dosimetry & SSDE
• Example 1: Multi-Slice Helical Dosimetry and Effective Dose Calculation

A medical physicist conducts an annual accreditation survey of an adult helical abdomen/pelvis protocol on a 64-slice scanner ( N = 64 , T = 0.625 mm N = 64, T = 0.625\text{ mm} N = 64 , T = 0.625 mm , total beam collimation N ⋅ T = 40.0 mm N \cdot T = 40.0\text{ mm} N ⋅ T = 40.0 mm ). Measurements in the 32 cm 32\text{ cm} 32 cm PMMA body phantom yield:

Central hole: C T D I 100 , center = 6.4 mGy CTDI_{100,\text{center}} = 6.4\text{ mGy} CT D I 100 , center ​ = 6.4 mGy

Peripheral holes: C T D I 12 = 14.2 mGy , C T D I 3 = 13.8 mGy , C T D I 6 = 14.4 mGy , C T D I 9 = 13.6 mGy CTDI_{12} = 14.2\text{ mGy}, CTDI_3 = 13.8\text{ mGy}, CTDI_6 = 14.4\text{ mGy}, CTDI_9 = 13.6\text{ mGy} CT D I 12 ​ = 14.2 mGy , CT D I 3 ​ = 13.8 mGy , CT D I 6 ​ = 14.4 mGy , CT D I 9 ​ = 13.6 mGy

Peripheral average: C T D I 100 , periphery = ( 14.2 + 13.8 + 14.4 + 13.6 ) / 4 = 14.0 mGy CTDI_{100,\text{periphery}} = (14.2 + 13.8 + 14.4 + 13.6) / 4 = 14.0\text{ mGy} CT D I 100 , periphery ​ = ( 14.2 + 13.8 + 14.4 + 13.6 ) /4 = 14.0 mGy

• Helical couch increment: Δ d = 36.0 mm/rotation \Delta d = 36.0\text{ mm/rotation} Δ d = 36.0 mm/rotation
• Scan length: L = 45 cm L = 45\text{ cm} L = 45 cm

Step 1: Calculate Weighted CTDI ( C T D I w CTDI_w CT D I w ​ ) C T D I w = 1 3 ( 6.4 mGy ) + 2 3 ( 14.0 mGy ) = 2.133 + 9.333 = 11.47 mGy CTDI_w = \frac{1}{3}(6.4\text{ mGy}) + \frac{2}{3}(14.0\text{ mGy}) = 2.133 + 9.333 = 11.47\text{ mGy} CT D I w ​ = 3 1 ​ ( 6.4 mGy ) + 3 2 ​ ( 14.0 mGy ) = 2.133 + 9.333 = 11.47 mGy

Step 2: Calculate Pitch and Volume CTDI ( C T D I vol CTDI_{\text{vol}} CT D I vol ​ ) pitch = Δ d N ⋅ T = 36.0 mm 40.0 mm = 0.90 \text{pitch} = \frac{\Delta d}{N \cdot T} = \frac{36.0\text{ mm}}{40.0\text{ mm}} = 0.90 pitch = N ⋅ T Δ d ​ = 40.0 mm 36.0 mm ​ = 0.90 C T D I vol = C T D I w pitch = 11.47 mGy 0.90 = 12.74 mGy CTDI_{\text{vol}} = \frac{CTDI_w}{\text{pitch}} = \frac{11.47\text{ mGy}}{0.90} = 12.74\text{ mGy} CT D I vol ​ = pitch CT D I w ​ ​ = 0.90 11.47 mGy ​ = 12.74 mGy Compliance Check : C T D I vol = 12.74 mGy CTDI_{\text{vol}} = 12.74\text{ mGy} CT D I vol ​ = 12.74 mGy is well below the ACR reference level of 25 mGy 25\text{ mGy} 25 mGy and pass/fail limit of 30 mGy 30\text{ mGy} 30 mGy .

Step 3: Calculate Dose-Length Product ( D L P DLP D L P ) and Effective Dose ( E E E ) D L P = C T D I vol × L = 12.74 mGy × 45 cm = 573.3 mGy ⋅ cm DLP = CTDI_{\text{vol}} \times L = 12.74\text{ mGy} \times 45\text{ cm} = 573.3\text{ mGy}\cdot\text{cm} D L P = CT D I vol ​ × L = 12.74 mGy × 45 cm = 573.3 mGy ⋅ cm Using the adult abdomen/pelvis coefficient k = 0.015 mSv / ( mGy ⋅ cm ) k = 0.015\text{ mSv}/(\text{mGy}\cdot\text{cm}) k = 0.015 mSv / ( mGy ⋅ cm ) : E = 0.015 × 573.3 = 8.60 mSv E = 0.015 \times 573.3 = 8.60\text{ mSv} E = 0.015 × 573.3 = 8.60 mSv

Example 2: AAPM Report 220 Water-Equivalent Diameter & SSDE

A 4-year-old child undergoes an emergency abdominal CT. The scanner reports a volume CTDI based on a 32 cm 32\text{ cm} 32 cm body phantom of C T D I vol = 5.2 mGy CTDI_{\text{vol}} = 5.2\text{ mGy} CT D I vol ​ = 5.2 mGy .

From the mid-slice axial CT contour, the automated radiation tracking software calculates:

Contour cross-sectional area: A R O I = 165 cm 2 A_{ROI} = 165\text{ cm}^2 A RO I ​ = 165 cm 2

Mean CT number of the cross section: CT ‾ R O I = − 35 HU \overline{\text{CT}}_{ROI} = -35\text{ HU} CT RO I ​ = − 35 HU

Step 1: Calculate Water-Equivalent Area ( A w A_w A w ​ ) and Water-Equivalent Diameter ( D w D_w D w ​ ) A w = [ − 35 1000 + 1 ] × 165 = [ 0.965 ] × 165 = 159.2 cm 2 A_w = \left[ \frac{-35}{1000} + 1 \right] \times 165 = [0.965] \times 165 = 159.2\text{ cm}^2 A w ​ = [ 1000 − 35 ​ + 1 ] × 165 = [ 0.965 ] × 165 = 159.2 cm 2 D w = 2 A w π = 2 159.2 3.1416 = 2 50.68 = 2 × 7.12 = 14.24 cm D_w = 2 \sqrt{\frac{A_w}{\pi}} = 2 \sqrt{\frac{159.2}{3.1416}} = 2 \sqrt{50.68} = 2 \times 7.12 = 14.24\text{ cm} D w ​ = 2 π A w ​ ​ ​ = 2 3.1416 159.2 ​ ​ = 2 50.68 ​ = 2 × 7.12 = 14.24 cm

Step 2: Determine Size Conversion Factor ( f size f_{\text{size}} f size ​ ) and SSDE Using the AAPM Report 204/220 parameterization for a 32 cm 32\text{ cm} 32 cm reference phantom ( f size = a ⋅ e − b ⋅ D w f_{\text{size}} = a \cdot e^{-b \cdot D_w} f size ​ = a ⋅ e − b ⋅ D w ​ , with a = 3.704369 , b = 0.03671937 a = 3.704369, b = 0.03671937 a = 3.704369 , b = 0.03671937 for 32 cm 32\text{ cm} 32 cm body): f size = 3.704369 × e − 0.03671937 × 14.24 = 3.704369 × e − 0.5229 = 3.704369 × 0.5928 = 2.20 f_{\text{size}} = 3.704369 \times e^{-0.03671937 \times 14.24} = 3.704369 \times e^{-0.5229} = 3.704369 \times 0.5928 = 2.20 f size ​ = 3.704369 × e − 0.03671937 × 14.24 = 3.704369 × e − 0.5229 = 3.704369 × 0.5928 = 2.20 S S D E = f size × C T D I vol = 2.20 × 5.2 mGy = 11.44 mGy SSDE = f_{\text{size}} \times CTDI_{\text{vol}} = 2.20 \times 5.2\text{ mGy} = 11.44\text{ mGy} SS D E = f size ​ × CT D I vol ​ = 2.20 × 5.2 mGy = 11.44 mGy Clinical Takeaway : While the console reported C T D I vol = 5.2 mGy CTDI_{\text{vol}} = 5.2\text{ mGy} CT D I vol ​ = 5.2 mGy , the child's true absorbed dose was 11.44 mGy 11.44\text{ mGy} 11.44 mGy ( 120 % 120\% 120% higher) due to the reduced attenuation of the small pediatric torso.

Comparison Table & Clinical Pitfalls

Dosimetric Parameter Mathematical Definition Standard Units Physical Significance Key Clinical Limitation

C T D I 100 CTDI_{100} CT D I 100 ​ 1 N ⋅ T ∫ − 50 + 50 D ( z ) d z \frac{1}{N \cdot T} \int_{-50}^{+50} D(z) \, dz N ⋅ T 1 ​ ∫ − 50 + 50 ​ D ( z ) d z mGy \text{mGy} mGy Single-rotation dose along a 100 mm pencil chamber Ignores helical pitch and radial dose variation

C T D I w CTDI_w CT D I w ​ 1 3 C T D I center + 2 3 C T D I periphery \frac{1}{3} CTDI_{\text{center}} + \frac{2}{3} CTDI_{\text{periphery}} 3 1 ​ CT D I center ​ + 3 2 ​ CT D I periphery ​ mGy \text{mGy} mGy Radial average dose in PMMA phantom Ignores helical pitch and table motion

C T D I vol CTDI_{\text{vol}} CT D I vol ​ C T D I w / pitch CTDI_w / \text{pitch} CT D I w ​ / pitch mGy \text{mGy} mGy Standardized scanner radiation output per unit volume Calibrated to rigid PMMA; not patient dose

D L P DLP D L P C T D I vol × L CTDI_{\text{vol}} \times L CT D I vol ​ × L mGy ⋅ cm \text{mGy}\cdot\text{cm} mGy ⋅ cm Total radiation energy imparted along scan length Excludes patient size and individual organ risks

S S D E SSDE SS D E f size × C T D I vol f_{\text{size}} \times CTDI_{\text{vol}} f size ​ × CT D I vol ​ mGy \text{mGy} mGy Patient-absorbed dose adjusted for body habitus ( D w D_w D w ​ ) Regional cross-sectional metric; not whole-body

Effective Dose ( E E E ) k × D L P k \times DLP k × D L P mSv \text{mSv} mSv Population stochastic radiation detriment Normalized model; invalid for individual risk projection

Clinical Practice Pitfalls

Conflating C T D I vol CTDI_{\text{vol}} CT D I vol ​ with Individual Patient Dose : Quoting C T D I vol CTDI_{\text{vol}} CT D I vol ​ to patients or clinicians misrepresents actual absorbed dose, especially in pediatric and bariatric cohorts.

Patient Off-Centering in the Gantry : Positioning a patient above or below isocenter causes erroneous magnification on scout radiographs, causing ATCM algorithms to miscalculate tube current and degrading bow-tie filter compensation (escalating surface dose and introducing artifacts).

Neglecting Lung Attenuation in Thoracic SSDE : Utilizing geometric diameters ( A P + L A T AP + LAT A P + L A T ) rather than water-equivalent diameter ( D w D_w D w ​ ) in chest imaging substantially overestimates tissue attenuation and underestimates SSDE.

Misinterpreting Vendor ATCM Parameter Directionality : Increasing the noise parameter reduces radiation dose on GE systems (Noise Index), whereas it increases dose on Siemens systems (Quality Reference mAs).