CT Patient Centering: Dose and Image Quality

Centering the patient at the scanner's axis of rotation is one of the simplest, cheapest, and most overlooked dose-reduction actions in CT. When a patient sits below isocenter—which is the most common error—the bowtie beam-shaping filter and the localizer-driven automatic exposure control both push in the wrong direction, raising surface dose while degrading image noise. Correct centering restores the dose-and-noise advantage the scanner was engineered to deliver, at no cost in hardware.¹²

This article explains the physics of why centering matters, quantifies the dose and noise penalty using published phantom and clinical data, and gives a practical positioning, QC, and workflow checklist that a diagnostic medical physicist can fold into a facility's CT program.

Introduction

Modern CT dose optimization is usually framed around protocol parameters: tube voltage, tube current modulation, pitch, reconstruction, and beam collimation. Those matter. But before any of them takes effect, the technologist makes one geometric decision—how high to raise the table—that quietly rescales the whole exposure. That decision determines whether the patient is at isocenter, where the scanner's beam-shaping filter and dose-efficiency assumptions hold, or displaced from it, where they no longer do.¹

The effect is not subtle. Controlled phantom studies show that a few centimeters of vertical offset can change surface dose by tens of percent and measurably increase image noise.¹² Retrospective reviews of clinical positioning show that a large fraction of real patients are consistently placed below isocenter.¹²³ Because the error is systematic and repeated on every patient, its cumulative impact on population dose and image quality is significant—yet it is invisible in a protocol printout.

For facilities that already invest in CT protocol optimization, tube current modulation, and dose metrics tracking, centering is the geometric prerequisite that makes those investments pay off. DRPS addresses it as part of CT physics testing and medical physics consulting across Florida, Maryland, Virginia, Washington DC, California, Nevada, and our other service areas.

Topic Explanation

What "centering" actually means

Patient centering in CT refers to placing the anatomic region of interest at the scanner's axis of rotation—the isocenter—by adjusting table height (and, secondarily, lateral position). The gantry rotates the x-ray tube and detector around this axis. The scanner's dose-shaping and exposure-control systems are all calibrated on the assumption that the thickest, most attenuating part of the patient sits at that axis.

Two scanner subsystems make centering matter:

• The bowtie (beam-shaping) filter. A physical filter mounted at the tube port that is thin in the middle and thick at the edges. It is designed so that the beam is most intense along the central ray—where it must penetrate the thickest part of a centered patient—and progressively attenuated toward the periphery, where the patient is thinner. This equalizes the fluence reaching the detector, reduces dynamic-range demands, cuts scatter, and lowers peripheral skin dose. Its shape only "fits" a patient centered at isocenter.¹
• Automatic exposure control (AEC). The tube-current-modulation system estimates patient attenuation—usually from one or two localizer radiographs (also called scout or topogram)—and adjusts tube current to hold a target image quality. Its size estimate is only accurate if the patient is centered when the localizer is acquired.²

When the patient is off-center, both systems operate on a false premise.

Why below-isocenter is the common failure mode

Gravity, table design, and habit conspire to place patients too low. Technologists frequently set table height by eye or to a fixed comfortable working level rather than to the anatomic midline, and larger or less mobile patients tend to settle low. Multiple studies quantify this: in one landmark analysis, 46% of adult body patients were miscentered by 20–60 mm, with a mean position roughly 23 mm below isocenter; clinical chest-CT reviews reported median offsets of about 25–35 mm below isocenter, with the effect more pronounced in smaller pediatric patients.¹²³ Because the error is directional and repeated, it is a systematic bias, not random noise.

Key Technical Principles

The bowtie-filter mismatch

When a patient drops below isocenter, the anterior surface moves toward the tube (on scanners where the tube passes over/under the patient during rotation) and into a region where the bowtie filter is thinner than intended for that path. The consequence is a higher entrance fluence over part of the patient surface and a loss of the dose-equalizing benefit the filter was designed to provide. The net effect measured at the surface of a body-sized phantom is an increase in peak surface dose as offset grows, with a simultaneous rise in image noise because the fluence reaching the detector is no longer well matched across the field of view.¹

The published phantom relationship is approximately monotonic with offset:

Values are representative measurements from controlled phantom work on a body-sized phantom and depend on scanner, bowtie selection, and phantom size; they are not universal constants.¹ The key qualitative point holds broadly: miscentering raises surface dose and degrades noise at the same time—there is no favorable trade-off.

The localizer / AEC magnification error

The second mechanism is geometric. A localizer radiograph is a projection: the x-ray tube is on one side of the patient and the detector on the other. A structure at isocenter is magnified by

where $\text{SID}$ is the source-to-image-receptor distance and $\text{SAD}$ is the source-to-axis (isocenter) distance. If the patient's mid-plane is displaced a distance $\Delta$ toward the tube, the source-to-object distance shrinks to $\text{SAD}-\Delta$ and the magnification grows to

so the apparent width the AEC extracts from the localizer inflates by the factor

Using representative CT geometry (source-to-isocenter distances of roughly 54–63 cm are typical across vendors), take $\text{SAD}=60$ cm, $\text{SID}=100$ cm, and a $\Delta =6$ cm displacement toward the tube:

The AEC therefore "sees" a patient about 11% wider than reality. Because tube current is escalated to compensate for the attenuation implied by apparent size, this inflated size systematically raises tube current—and dose—on top of the bowtie penalty. On many scanners the localizer tube passes beneath the table, so a patient placed too low sits closer to the tube and is over-magnified, which is exactly consistent with the empirical finding that most patients are both too low and over-dosed.¹² The specific numbers are vendor- and geometry-dependent, but the mechanism is general: any vertical offset makes the single-projection size estimate wrong, and AEC faithfully acts on the wrong number.

A lateral or dual/orthogonal localizer is far less sensitive to vertical offset, which is why using a lateral projection for the AEC size estimate reduces (though does not eliminate) the centering-driven current error.² The bowtie surface-dose penalty remains regardless of localizer strategy.

Interaction with SSDE and dose metrics

Because miscentering distorts the apparent size, it also distorts anything derived from apparent size. The size-specific dose estimate (SSDE) uses a measured effective diameter to convert ${\text{CTDI}}_{vol}$ into a patient-specific dose; if the effective diameter is taken from an off-center localizer, both the water-equivalent diameter and the SSDE conversion inherit the geometric error. Likewise, dashboards that trend ${\text{CTDI}}_{vol}$ and DLP across a fleet will see centering show up as unexplained scatter in dose and noise—one reason large multi-facility analyses identify patient centering among the influential factors driving variability in image quality and dose.⁴

Clinical Impact

The clinical cost of chronic miscentering is a steady, avoidable excess in patient dose paired with worse image quality—the opposite of what optimization is supposed to achieve. Several groups have quantified organ-level consequences. In a pediatric anthropomorphic phantom, vertical offsets of up to ±6 cm changed organ doses by up to about 22–34% depending on the organ and body region, with an inverse relationship between position and image noise; the head was driven mainly by the bowtie filter while the thoracoabdominal region was affected by both the bowtie filter and AEC.³ Superficial radiosensitive organs—breast, thyroid, eye lens—are especially exposed to the surface-dose penalty because they sit near the entrance surface where the bowtie mismatch is largest.³

Children matter disproportionately. Smaller patients are both more likely to be miscentered and more sensitive to a given absolute offset, because a 3 cm error is a larger fraction of a small patient's diameter.²³ For pediatric protocols already engineered toward dose minimization, poor centering can silently erase the benefit of a carefully lowered technique.

The image-quality side is just as real. Increased noise from miscentering can prompt technologists to raise technique or repeat marginal series, compounding dose. When noise is worst exactly where anatomy is thickest and centering is poorest, low-contrast detectability suffers where it is often most needed.

Practical Optimization Tips

Centering is a workflow and QC problem more than a hardware problem. The following steps address it at both levels.

1. Make isocenter the explicit target, not table comfort

Train technologists to set table height so the mid-coronal plane of the anatomy of interest is at isocenter—using the scanner's positioning laser or centering light on the anatomic midline—rather than to a habitual working height. For the thorax and abdomen, the mid-axillary line is a practical landmark; for the head, the external auditory meatus / orbitomeatal reference.

2. Verify with the localizer before scanning

The localizer is a free check. A quick look confirms the anatomy is vertically centered in the display field before the helical acquisition starts. Building a "glance at the localizer for centering" habit catches gross errors before dose is delivered.

3. Prefer localizer strategies that are robust to offset

Where the scanner and protocol allow, use a lateral or dual/orthogonal localizer for AEC size estimation so that residual vertical offset has less influence on the tube-current calculation.² Confirm the AEC configuration with the vendor and your physicist so you know which projection drives modulation.

4. Commission and check automated positioning

Camera- or depth-sensor-based auto-positioning systems (offered by major vendors) estimate the body contour and set or recommend table height toward isocenter, reducing operator variability. These systems should be commissioned at acceptance, verified against the scanner's true isocenter, and rechecked during annual physics testing—an automated system that is miscalibrated simply automates a systematic error.

5. Audit centering as a QC metric

Periodically review a sample of localizers or use available positioning-analytics tools to detect systematic low positioning. Fold the finding into technologist feedback and into image-quality/dose reviews. Trend it the way you trend ${\text{CTDI}}_{vol}$ outliers.

Common pitfalls to avoid

• Assuming AEC compensates for centering. It does not correct the bowtie penalty and can be actively misled by an off-center localizer.¹²
• Setting table height by habit. A fixed comfortable height is not isocenter for most patients.
• Ignoring pediatric sensitivity. Small patients are the most miscentered and the most affected.³
• Trusting an unverified auto-positioning system. Automation must be commissioned and periodically checked.
• Treating centering as "just positioning." It is a dose-and-image-quality control point that belongs in the physics QC program.

Regulatory Considerations

Patient centering sits inside the broader diagnostic-physics performance-monitoring framework rather than in a single rule, but it is squarely a qualified-medical-physicist and QC responsibility. CT is regulated as a radiation-producing machine under state radiation-control programs (X-ray machines are FDA plus state/Agreement-State regulated), and dose-and-image-quality performance is governed through accreditation and medical-physics standards.

• ACR–AAPM Technical Standard for Diagnostic Medical Physics Performance Monitoring of CT (2022 revision). Establishes the qualified medical physicist's role in performance monitoring, including image quality, dose, and the acceptance and annual evaluation of CT systems—the natural home for verifying isocenter, positioning aids, and any automated centering system.⁵
• ACR CT Quality Control Manual and CT Accreditation Program. Provide the routine QC structure (technologist QC, medical-physicist testing) within which centering checks, localizer review, and dose/noise consistency are monitored.⁶
• IEC 60601-2-44. The particular standard for CT equipment safety and performance, including automatic exposure control behavior that depends on correct patient positioning.⁷
• Size-specific dose guidance (AAPM Report No. 204, SSDE). Because SSDE depends on a measured effective diameter, centering directly affects the accuracy of size-specific dose reporting—reinforcing centering as a basic element of dose management and ALARA in CT.⁸

For facilities pursuing or maintaining accreditation, centering should be documented as part of the CT QC program and technologist competency, and confirmed at acceptance and annually by a qualified medical physicist. DRPS integrates this into CT physics testing, accreditation support, and radiation safety training. Requirements vary by state radiation-control authority, so confirm specifics with the authority having jurisdiction.

Frequently Asked Questions (FAQs)

Why does patient centering matter so much in CT?

CT scanners use a bowtie (beam-shaping) filter and automatic exposure control that are both optimized for a patient centered at the scanner's axis of rotation (isocenter). When the patient is off-center, the thin part of the bowtie filter no longer lines up with the thickest part of the patient, so surface dose rises and dose efficiency falls. If the localizer radiograph is used for automatic exposure control, off-centering also distorts the apparent patient size, which can drive tube current—and dose—up or down inappropriately.

Which direction do patients usually get miscentered?

Studies consistently find that most patients are positioned below isocenter—too low on the table. Toth and colleagues found 46% of adult body patients were miscentered 20–60 mm, with a mean position about 23 mm below isocenter. Being too low typically increases surface dose and, on many scanners, inflates the apparent patient size on the localizer, further raising tube current.

How much extra dose does miscentering cause?

It depends on patient size, offset, and scanner. In phantom work, a 3 cm and 6 cm vertical offset increased surface dose on a 32 cm CTDI phantom by about 18% and 41%, respectively, with image noise increasing about 6% and 22%. In clinical chest CT, adult phantom dose rose about 38% at the lowest table position and fell about 23% at the highest. These are representative figures, not fixed constants.

Does automatic exposure control fix miscentering?

No. Automatic exposure control (AEC) manages tube current for a target image quality, but it does not correct the bowtie-filter mismatch that raises surface dose. Worse, when AEC is driven by a single localizer acquired with the patient off-center, the apparent patient size is distorted, so AEC can systematically over- or under-expose. Proper centering is required for AEC to work as intended.

Can a lateral localizer or dual localizers reduce the problem?

A lateral (or dual/orthogonal) localizer reduces the sensitivity of AEC size estimation to vertical offset, because a lateral projection is far less affected by up/down displacement than a single anteroposterior projection. Many facilities use two orthogonal localizers or a lateral localizer specifically to make AEC more robust to residual centering error. It still does not remove the bowtie-filter surface-dose penalty.

How do 3D camera auto-positioning systems help?

Camera- or depth-sensor-based auto-positioning systems estimate the patient's body contour and recommend or automatically set table height to place the patient at isocenter, reducing operator-to-operator variability. They can improve centering consistency, but they should be commissioned and periodically checked by a medical physicist, and technologists still need training and a manual verification habit.

What should a CT QC program check regarding centering?

Centering should be addressed in both routine QC and technologist training: verify table-height indicators and positioning lasers against isocenter, review a sample of localizers for systematic low positioning, include centering in image-quality and dose reviews, and confirm any automated positioning system during acceptance and annual physics testing. A qualified medical physicist can build this into the ACR–AAPM performance-monitoring framework.

Key Takeaways

• Centering is a dose-and-image-quality control point, not just comfort. The bowtie filter and AEC are both calibrated for a patient at isocenter.¹²
• Below-isocenter is the common, systematic error. Nearly half of adult body patients in one study were low by 20–60 mm; pediatric patients are affected more.¹²³
• The penalty is measurable. A 6 cm offset raised surface dose about 41% and noise about 22% on a body phantom; clinical chest CT varied roughly −23% to +38% across table height.¹²
• AEC does not save you. It cannot fix the bowtie mismatch and can be misled by an off-center single localizer; lateral/dual localizers help.²
• Automation helps but must be verified. Camera-based auto-positioning improves consistency only if commissioned and checked by a physicist.
• Fold it into QC. Verify isocenter and positioning aids, audit localizers, and address centering in accreditation and annual physics testing.⁵⁶

Conclusion

Patient centering is the geometric foundation on which every other CT dose-optimization strategy rests. The bowtie beam-shaping filter and automatic exposure control are engineered around a patient at isocenter; when the patient sits below it—as most do—surface dose rises and image noise worsens simultaneously, and the localizer-based size estimate that drives AEC can compound the error. None of this requires new hardware to fix. It requires training technologists to target isocenter, choosing localizer strategies that are robust to residual offset, commissioning and checking any automated positioning system, and treating centering as a monitored QC metric. Done consistently, correct centering recovers the dose efficiency and image quality the scanner was designed to deliver.

How DRPS Can Help

Diagnostic Radiation Physics Services helps CT facilities turn centering from an invisible habit into a controlled, documented part of the physics program. This can include acceptance and annual CT physics testing that verifies isocenter and positioning aids, commissioning of camera-based auto-positioning systems, review of clinical localizers for systematic miscentering, technologist training, and integration of centering into dose and image-quality monitoring and accreditation support.

DRPS supports imaging facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware, with work performed by board-certified medical physicists.

Related Resources

• CT protocol optimization
• CT tube current modulation
• Size-specific dose estimate (SSDE) in CT
• CTDIvol and DLP dose metrics
• Pediatric CT dose optimization
• CT physics testing
• Medical physicist consulting

References

1. Toth T, Ge Z, Daly MP. The influence of patient centering on CT dose and image noise. Medical Physics. 2007;34(7):3093-3101. doi:10.1118/1.2748113. doi.org
2. Kaasalainen T, Palmu K, Reijonen V, Kortesniemi M. Effect of patient centering on patient dose and image noise in chest CT. AJR American Journal of Roentgenology. 2014;203(1):123-130. doi:10.2214/AJR.13.12028. doi.org
3. Euler A, Saltybaeva N, Alkadhi H. How patient off-centering impacts organ dose and image noise in pediatric head and thoracoabdominal CT. European Radiology. 2019;29(12):6790-6793. doi:10.1007/s00330-019-06330-5. doi.org
4. Smith TB, Zhang S, Erkanli A, Frush D, Samei E. Variability in image quality and radiation dose within and across 97 medical facilities. Journal of Medical Imaging. 2021;8(5):052105. doi:10.1117/1.JMI.8.5.052105. doi.org
5. American College of Radiology, American Association of Physicists in Medicine. ACR–AAPM Technical Standard for Diagnostic Medical Physics Performance Monitoring of Computed Tomography (CT) Equipment. Revised 2022. acr.org
6. American College of Radiology. Computed Tomography Quality Control Manual and CT Accreditation Program requirements. acr.org
7. International Electrotechnical Commission. IEC 60601-2-44: Medical electrical equipment — Particular requirements for the basic safety and essential performance of X-ray equipment for computed tomography. Geneva: IEC. iec.ch
8. American Association of Physicists in Medicine. Size-Specific Dose Estimates (SSDE) in Pediatric and Adult Body CT Examinations. AAPM Report No. 204. College Park, MD: AAPM; 2011. aapm.org