Beyond CTDI100: Wide-Beam CT Dose Metrics

Introduction

The computed tomography dose index that every CT scanner displays was designed for a beam a centimeter wide, and a 100 mm measuring chamber cannot capture the dose spread of a modern wide-beam or stationary cone-beam acquisition. As detectors grew from a single slice to banks covering 80, 120, or 160 mm of anatomy in one rotation, the measurement at the heart of CT dosimetry quietly drifted away from the thing it was meant to describe.

CTDI100 — the dose index measured with a 100 mm pencil ionization chamber — has been the workhorse of CT dosimetry for decades. It is reproducible, standardized, and easy to measure, and it underpins the CTDIvol and dose-length product values that scanners report, that registries track, and that physicists use to compare protocols. ^{1, 10} But the dose profile from a CT rotation is not confined to the beam: it has long, low-amplitude scatter tails that reach well past the ends of a 100 mm chamber. ⁵ When those tails are truncated, the measured index falls short of the dose actually accumulated. ^{1, 3}

AAPM Report No. 111 — the report of Task Group 111 — reframed the measurement around that physics. ¹⁰ Instead of integrating a truncated profile, it measures the equilibrium dose: the dose at a fixed point in a long phantom when the accumulated dose has stopped rising as the scan gets longer. This guide explains why CTDI100 undercounts, what the equilibrium approach measures instead, where CTDIvol stays valid and where it breaks, and how to connect all of this to size-specific dosimetry and defensible QC. DRPS provides this kind of analysis as part of its CT physics testing and medical physics consulting services across Florida, Maryland, Virginia, Washington DC, California, and Nevada.

Topic Explanation

What CTDI100 actually measures

CTDI100 is defined as the integral of the single-rotation dose profile $D (z)$ along the scanner's $z$ -axis, measured over a 100 mm length and normalized by the nominal beam width:

CTDI_{100} = \frac{1}{n T} \int_{- 50 mm}^{+ 50 mm} D (z) d z

Here $n T$ is the total nominal collimated beam width (number of channels $n$ times slice thickness $T$ ). The measurement is made in standardized polymethylmethacrylate (PMMA) phantoms — 16 cm diameter for "head," 32 cm for "body" — at the center and periphery, then combined into a weighted index:

CTDI_{w} = \frac{1}{3} CTDI_{100, center} + \frac{2}{3} CTDI_{100, periphery}

Dividing by pitch gives $CTDI_{v o l}$ , and multiplying by scan length gives the dose-length product (DLP). These quantities are formally defined for CT scanners in IEC 60601-2-44, the standard that governs what the console displays. ¹⁴

When CT beams were 10 mm wide or less, the 100 mm chamber comfortably spanned the whole dose profile, and CTDI100 was an excellent descriptor. The problem is not that CTDI100 was ever wrong — it is that the hardware moved and the measurement did not.

The scatter tail that escapes the chamber

The dose deposited at a point in a cylindrical phantom during one rotation is not just the primary beam. Scattered photons spread the dose along $z$ in long, low-amplitude tails. Monte Carlo work by Boone quantified these dose spread functions: for a body-sized phantom at 120 kVp, the central dose spread function has a full width at tenth maximum of roughly 160 mm, rising to about 250 mm for a 50 cm phantom. ⁵ The scatter-to-primary dose ratio at the center of a 32 cm phantom reaches about 12.5. ⁵ A 100 mm chamber simply cannot be long enough to collect a tail that wide.

This is the key insight, and it is easy to get backward. The trouble is not primarily that beams got wider. Boone's efficiency analysis showed that CTDI100 collects only about 63 percent of the dose at the center of a 32 cm body phantom and about 88 percent at the periphery at 120 kVp — and that this efficiency drops by only about one percentage point going from a 10 mm to a 40 mm beam. ³ In other words, CTDI100 was already truncating the scatter tail even for the narrow beams it was designed for; widening the beam barely changes the collection efficiency. The deficit was there all along; wide beams and new acquisition modes simply made it impossible to ignore. ³

For more on how noise, resolution, and dose interact in CT image quality, see our companion articles on the CT noise power spectrum and task-based image quality and CTDI measurement and CT dose QC.

Key Technical Principles

Equilibrium dose: the AAPM Report 111 approach

AAPM Report 111 proposed measuring dose differently. ¹⁰ Rather than integrating a single truncated profile, place a small ion chamber at a fixed point in a long phantom and translate the phantom through a scan. As the scan length $L$ increases, the accumulated dose at the chamber climbs — each added rotation contributes primary dose plus a scatter tail reaching back to the measurement point — until it plateaus. That plateau is the equilibrium dose, $D_{e q}$ .

Formally, the equilibrium dose equals the complete dose-profile integral spread over the scan increment $b$ (the table feed per rotation):

D_{e q} = \frac{1}{b} \int_{- \infty}^{+ \infty} f (z) d z

where $f (z)$ is the full axial dose profile, scatter tails included. Because the integration is over the entire profile rather than a 100 mm window, nothing is truncated.

The rise toward that plateau is described by an approach-to-equilibrium function $H (L)$ , the fraction of the equilibrium dose reached for a scan of length $L$ :

H (L) = \frac{D _{L} ( 0 )}{D _{e q}} = 1 - e^{- L / d_{e q}}

where $d_{e q}$ is a characteristic equilibrium length on the order of hundreds of millimeters in body tissue. Dixon and Boone showed that this same $H$ function, and a single equilibrium-dose parameter $A_{e q}$ that is independent of the collimator aperture, unifies conventional helical and axial scanning with stationary cone-beam acquisition — the dose on the central ray of a wide cone beam, $f (0)$ , is numerically equal to the dose CTDI predicts for an axial series at the same point. ^{6, 7} The practical payoff is that one short-chamber measurement can characterize dose across scan modes. ⁷

How much does the 100 mm window miss?

The experimental numbers are consistent across independent groups. Dixon's original derivation predicted that CTDI100 underestimates the equilibrium dose by about 20 percent at the center of the body phantom and about 10 percent at the center of the head phantom, even for a 20 mm beam. ¹ Mori and colleagues, measuring on a 256-channel prototype, found that a 100 mm integration captured only 76 percent of the dose-profile integral at a 20 mm beam width and 60 percent at a 138 mm beam width in a 32 cm body phantom, and that an integration length of at least 300 mm was needed to recover more than 90 percent. ² Dixon and Ballard measured the ratio directly on clinical GE multidetector scanners at 120 kVp in a 32 cm body phantom: the equilibrium dose was 1.75 times the central CTDI100 and 1.22 times the peripheral CTDI100. ⁴

Metric	Narrow beam (≈20 mm)	Wide beam (≈138 mm)	Source
Fraction of dose-profile integral captured by a 100 mm integration (32 cm body phantom)	≈76%	≈60%	Mori 2005 ²
Integration length needed to capture more than 90% of the profile	about 300 mm	about 300 mm	Mori 2005 ²
CTDI100 collection efficiency, body-phantom center (120 kVp)	≈63%	≈62% (≈1% lower than 10 mm)	Boone 2007 ³
Equilibrium dose relative to central CTDI100 (GE MDCT, 120 kVp, 32 cm)	≈1.75×	≈1.75×	Dixon and Ballard 2007 ⁴

A practical conversion factor from a 100 mm to a 300 mm integration, measured on a 256-slice scanner, ranges from about 1.5 to 2.0 and rises with both beam width and beam energy. ⁹

A worked example

Suppose a physicist measures a central CTDI100 of 20 mGy and a peripheral CTDI100 of 30 mGy in the 32 cm body phantom (illustrative inputs). The weighted index is

CTDI_{w} = \frac{1}{3} (20) + \frac{2}{3} (30) = 26.7 mGy

But the dose actually accumulating at the center of a long body scan is closer to the equilibrium value. Applying the measured central ratio from Dixon and Ballard: ⁴

D_{e q, center} = 1.75 \times CTDI_{100, center} = 1.75 \times 20 mGy = 35 mGy

The central equilibrium dose is about 75 percent higher than the central CTDI100 that would be measured with the pencil chamber. The ratios here (1.75 and 1.22) are the measured quantities; the 20 and 30 mGy are illustrative stand-ins for a facility's own measurement. The point is structural: for a long body scan, the pencil-chamber index systematically understates the central dose, and the gap is large enough to matter.

Clinical Impact

Where CTDIvol stays valid

None of this means CTDIvol should be discarded. For conventional moving-table helical and axial scanning, CTDIvol remains a standardized, reproducible index, it is what the scanner displays under IEC rules, and it is the backbone of dose-tracking registries and protocol comparison. ¹⁴ Its role as a relative, comparative index — is this protocol higher or lower than that one, is this scanner drifting over time — is unaffected by the equilibrium-dose discussion. CTDIvol was never meant to be the organ dose a patient receives, and understanding its limits is what lets a physicist use it correctly rather than abandon it.

Where CTDIvol breaks down

The failure mode is specific and increasingly common: stationary-table wide cone-beam acquisition (sometimes called SCBCT), along with perfusion studies and CT fluoroscopy. ^{6, 8} When the table does not move, the CTDI paradigm — which assumes a series of identical, equally spaced dose profiles — simply does not apply. Dixon and Boone documented scanner-reported CTDIvol values for stationary-phantom protocols that were more than 300 percent higher than the actual dose to the phantom, because the console applies a moving-table formula (with its extra-scatter and overbeaming corrections) to an acquisition where neither correction belongs. ⁸ A dose-tracking program that ingests those values uncritically will flag "bad data" that is really a metric mismatch. ⁸ For these acquisitions, the appropriate descriptor is the peak point dose $f (0)$ or the equilibrium dose, not an integral index. ⁶

The patient-size dimension

A separate but related gap is that a phantom-referenced index is not a patient dose. The size-specific dose estimate (SSDE) corrects CTDIvol for patient size using conversion factors keyed to the patient's effective diameter (AAPM Report 204) or, more robustly, to the attenuation-based water-equivalent diameter (AAPM Report 220). ^{11, 12} The CT Dose Index Registry analysis by Kanal and colleagues built U.S. diagnostic reference levels and achievable doses for the ten most common adult CT examinations as a function of patient size, reporting CTDIvol, DLP, and SSDE together — a reminder that size correction and dose-profile completeness are distinct corrections that belong in the same conversation. ¹⁵ For the broader size-correction topic, see our article on the size-specific dose estimate in CT.

Practical Optimization Tips

A practical approach to CT dose metrics on a modern scanner:

1. Know which metric your acquisition demands

For routine moving-table helical and axial scanning, use CTDIvol and DLP as comparative indices and for registry tracking.
For stationary wide cone-beam, perfusion, and CT fluoroscopy, treat scanner-reported CTDIvol with caution and prefer a point-dose or equilibrium-dose description. ^{6, 8}

2. Watch for anomalous displayed values

A CTDIvol that is implausibly high for a stationary acquisition is a classic metric mismatch, not necessarily a dose problem. ⁸
In a dose registry, outliers from perfusion or stationary protocols often reflect the metric, not the technique.

3. Measure the equilibrium dose where it matters

For wide-beam commissioning, a small-chamber equilibrium measurement in a long phantom characterizes dose more completely than a pencil-chamber CTDI100 alone. ^{4, 10}
One short-chamber measurement can, via the $H (L)$ framework, describe both conventional and stationary acquisitions. ⁷

4. Keep size correction separate and explicit

Apply SSDE using effective diameter or, preferably, water-equivalent diameter — do not conflate it with the equilibrium-dose correction. ^{11, 12}

5. Document the method, not just the number

Record the phantom, chamber, beam width, and integration length behind every dose value so results are interpretable and comparable over time.

Common pitfalls to avoid

Assuming CTDI100 captures the whole profile. It truncates the scatter tail even for narrow beams. ³
Blaming beam width alone. Collection efficiency changes only slightly with width; the scatter tail is the real driver. ³
Trusting a displayed CTDIvol for a stationary acquisition. It can overstate the true phantom dose by a wide margin. ⁸
Treating CTDIvol or SSDE as organ dose. They are indices and size-corrected indices, not the dose to any specific organ.
Discarding CTDIvol entirely. It remains a valid comparative and tracking index for conventional scanning. ¹⁴

Regulatory Considerations

CT dose metrics live at the intersection of an international equipment standard, national dose-reference work, and accreditation. The quantities themselves — CTDIvol and DLP — are defined for CT scanners in IEC 60601-2-44; the consolidated third edition (Edition 3.2, 2016) was the version in force as of this writing, and it governs what the console displays and how. ¹⁴ AAPM Report 111 provides the equilibrium-dose methodology that addresses the pencil-chamber's limits, and the companion ICRU Report 87 and AAPM Report 200 describe the long dosimetry phantom and the quantities used to implement it. ^{10, 13}

For patient-size correction and benchmarking, AAPM Report 204 and AAPM Report 220 define SSDE and the water-equivalent diameter, and the national CT Dose Index Registry analysis provides U.S. diagnostic reference levels and achievable doses by examination and patient size. ^{11, 12, 15} Diagnostic reference levels are not dose limits; they are comparison benchmarks that flag protocols for review when a facility's typical doses sit above the reference percentile.

In the United States, CT scanners are X-ray-producing devices regulated by the FDA and by state radiation-control programs, and routine physics performance evaluation is expected under accreditation (for example, ACR CT accreditation) and state rules. A defensible CT dose program documents which metric applies to each protocol, confirms that displayed values are trustworthy, and connects dose indices to size correction and reference levels. DRPS supports this through CT physics testing, accreditation support, and medical physics consulting. For the regulatory landscape across jurisdictions, see our overview of NRC, Agreement States, and FDA jurisdiction.

Frequently Asked Questions (FAQs)

What is CTDI100 and why was it created?

CTDI100 is the computed tomography dose index measured by integrating the single-rotation dose profile over a 100 mm long pencil ionization chamber in a PMMA phantom. It was designed when CT beams were a centimeter or less wide and a 100 mm integration captured essentially the entire dose profile, giving a convenient, reproducible index for comparing scanners and protocols.

Why does CTDI100 underestimate dose on modern scanners?

The dose profile from a CT rotation has long, low-amplitude scatter tails that extend far beyond 100 mm, so a 100 mm chamber misses part of the deposited dose even for narrow beams, and misses more as the phantom gets larger. The dominant problem is the un-captured scatter tail, not the beam width itself; a 100 mm integration can capture only about 60 percent of the dose-profile integral for a very wide beam in a body phantom.

What does AAPM Report 111 propose instead?

AAPM Report 111 measures the equilibrium dose using a small ion chamber held at a fixed point while the phantom is translated through a scan long enough for the accumulated dose at that point to stop rising. This equilibrium dose captures the full scatter contribution and applies uniformly to axial, helical, and wide cone-beam acquisitions, with or without table motion.

Is CTDIvol wrong, and should I stop using it?

CTDIvol remains a useful, standardized index for routine comparison and dose tracking for conventional helical and axial scanning, and it is what scanners display under IEC rules. It becomes unreliable for stationary-table wide cone-beam acquisitions, perfusion studies, and CT fluoroscopy, where a scanner-reported CTDIvol can be far higher than the true phantom dose. For those protocols a point-dose or equilibrium-dose description is more appropriate.

How is this different from SSDE?

The equilibrium-dose question is about how completely the measurement captures the dose profile. The size-specific dose estimate, or SSDE, is a separate correction that scales a phantom-referenced CTDIvol to a specific patient size using conversion factors keyed to effective diameter or water-equivalent diameter. The two address different gaps and are used together.

Does a wider detector automatically mean more dose to the patient?

No. Beam width by itself changes collection efficiency only slightly. What a wide beam changes is which dose metric is valid, and whether the acquisition is a moving-table helical scan or a stationary cone-beam rotation. The dose to the patient depends on technique, modulation, and anatomy, and must be judged with a metric appropriate to the acquisition mode.

When should a facility have a physicist review its CT dose metrics?

Review is warranted when commissioning a wide-detector or cone-beam-capable scanner, when adding perfusion or stationary-acquisition protocols, when displayed CTDIvol values look anomalous in a dose-tracking registry, and during annual performance evaluation and accreditation. A physicist can confirm which metric applies to each protocol and whether displayed values are trustworthy.

Key Takeaways

CTDI100 truncates the scatter tail. A 100 mm chamber misses part of the dose profile even for narrow beams, capturing only about 63 percent of the dose at the center of a body phantom. ³
Wide beams expose, but do not cause, the deficit. Collection efficiency changes by only about one percentage point from a 10 mm to a 40 mm beam; the scatter tail is the real driver. ³
The equilibrium dose fixes the measurement. AAPM Report 111 uses a small chamber in a long phantom to capture the full profile, and the equilibrium dose can be 1.75 times the central CTDI100 in a body phantom. ^{4, 10}
CTDIvol breaks for stationary cone-beam. For stationary-table, perfusion, and CT fluoroscopy acquisitions, displayed CTDIvol can overstate the true phantom dose by more than 300 percent; use a point-dose description instead. ⁸
CTDIvol still works for conventional scanning. As a comparative and tracking index for moving-table helical and axial scans, it remains valid and standardized. ¹⁴
Size correction is a separate step. SSDE scales a phantom index to patient size via effective or water-equivalent diameter — distinct from the equilibrium-dose correction. ^{11, 12}

Conclusion

The measurement at the center of CT dosimetry was built for a beam the width of a finger, and the hardware has long since outgrown it. The scatter tails that a 100 mm pencil chamber cannot reach were always there; wide detectors and stationary cone-beam acquisitions simply made the shortfall impossible to ignore. AAPM Report 111's equilibrium-dose method restores a measurement that matches the physics — capturing the full dose profile with a small chamber in a long phantom, and applying consistently across scan modes.

The practical lesson is not that CTDIvol is broken, but that it has a domain of validity. For conventional scanning it remains an excellent comparative index; for stationary cone-beam and perfusion it can mislead badly, and a point-dose or equilibrium description belongs in its place. A medical physicist who knows which metric applies to each protocol, who recognizes an anomalous displayed value for what it is, and who keeps size correction explicit will produce a CT dose program that is both accurate and defensible.

How DRPS Can Help

Diagnostic Radiation Physics Services helps CT facilities build dose programs that stand up to scrutiny. That work includes CT physics testing and acceptance testing for wide-detector and cone-beam-capable scanners, equilibrium-dose and CTDI measurements, protocol review, dose-registry interpretation, SSDE and water-equivalent-diameter implementation, accreditation support, and medical physics consulting — all performed by board-certified medical physicists.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware.

Knowing the limits of a dose metric is what makes the number trustworthy — and a trustworthy number is what protects patients and withstands an inspection.

Related Resources

References

Dixon RL. A new look at CT dose measurement: beyond CTDI. Medical Physics. 2003;30(6):1272-1280. doi:10.1118/1.1576952. doi.org
Mori S, Endo M, Nishizawa K, et al. Enlarged longitudinal dose profiles in cone-beam CT and the need for modified dosimetry. Medical Physics. 2005;32(4):1061-1069. doi:10.1118/1.1877852. doi.org
Boone JM. The trouble with CTDI100. Medical Physics. 2007;34(4):1364-1371. doi:10.1118/1.2713240. doi.org
Dixon RL, Ballard AC. Experimental validation of a versatile system of CT dosimetry using a conventional ion chamber: beyond CTDI100. Medical Physics. 2007;34(8):3399-3413. doi:10.1118/1.2757084. doi.org
Boone JM. Dose spread functions in computed tomography: a Monte Carlo study. Medical Physics. 2009;36(10):4547-4554. doi:10.1118/1.3223634. doi.org
Dixon RL, Boone JM. Cone beam CT dosimetry: a unified and self-consistent approach including all scan modalities—with or without phantom motion. Medical Physics. 2010;37(6):2703-2718. doi:10.1118/1.3395578. doi.org
Dixon RL, Boone JM. Analytical equations for CT dose profiles derived using a scatter kernel of Monte Carlo parentage with broad applicability to CT dosimetry problems. Medical Physics. 2011;38(7):4251-4264. doi:10.1118/1.3600518. doi.org
Dixon RL, Boone JM. Stationary table CT dosimetry and anomalous scanner-reported values of CTDIvol. Medical Physics. 2014;41(1):011907. doi:10.1118/1.4845075. doi.org
Mori S, Nishizawa K, Ohno M, Endo M. Conversion factor for CT dosimetry to assess patient dose using a 256-slice CT scanner. British Journal of Radiology. 2006;79(947):888-892. doi:10.1259/bjr/66519303. doi.org
American Association of Physicists in Medicine. AAPM Report No. 111: Comprehensive Methodology for the Evaluation of Radiation Dose in X-Ray Computed Tomography. Report of AAPM Task Group 111. 2010. aapm.org
American Association of Physicists in Medicine. AAPM Report No. 204: Size-Specific Dose Estimates (SSDE) in Pediatric and Adult Body CT Examinations. 2011. aapm.org
American Association of Physicists in Medicine. AAPM Report No. 220: Use of Water Equivalent Diameter for Calculating Patient Size and Size-Specific Dose Estimates (SSDE) in CT. 2014. aapm.org
International Commission on Radiation Units and Measurements. ICRU Report 87: Radiation Dose and Image-Quality Assessment in Computed Tomography. Journal of the ICRU. 2012;12(1). icru.org
International Electrotechnical Commission. IEC 60601-2-44: Medical electrical equipment — Part 2-44: Particular requirements for the basic safety and essential performance of X-ray equipment for computed tomography. Edition 3.2, 2016. iec.ch
Kanal KM, Butler PF, Sengupta D, Bhargavan-Chatfield M, Coombs LP, Morin RL. U.S. diagnostic reference levels and achievable doses for 10 adult CT examinations. Radiology. 2017;284(1):120-133. doi:10.1148/radiol.2017161911. doi.org