CT Patient Centering: Why It Changes Dose and Image Noise

You set the protocol, you picked the right noise level, and the images still look grainy in the back of the abdomen. One common cause is something that happens before the scan starts: where the patient sits in the gantry. CT patient centering decides how well the bowtie filter and tube current modulation can do their jobs, and a few centimeters of table height can change both dose and image noise.

This guide explains why centering matters, how it interacts with the scout, and what to check at the scanner on every patient.

Why isocenter matters: the bowtie filter

Every CT scanner has a shaped filter between the tube and the patient, usually called a bowtie filter. It sends the most X-rays through the middle of the scan field and fewer toward the edges, because the middle of a centered patient is the thickest part.

According to GE's Blueprint: A Guide to CT Radiation Dose Management, when the patient is off-center, the thickest part of the body no longer gets the full beam. Noise rises there, while thinner parts closer to the edge get more dose than they need. The same guide notes that being as little as 4 cm off-center could double the dose.

Siemens' SOMATOM Perspective CT Examination instructions (syngo CT VC40) describe the gantry laser light markers as showing the isocenter and the scan plane, and call for a monthly light-marker check, because a misaligned marker can make the exam unusable. If your lasers are off, your centering is off.

What the research shows

Miscentering is common, and it is usually low (patient too far down in the gantry):

• A 2012 World Journal of Radiology study of abdominal CT centering found that 81% of 395 patients were not correctly centered, and most of those were below isocenter. Image noise rose with off-centering, mostly in the posterior abdominal wall and liver.
• A Helsinki team's phantom and patient study in the American Journal of Roentgenology, summarized by Diagnostic Imaging in 2014, found that an adult phantom 6 cm below isocenter received about 38% more CTDIvol with a PA scout driving the automatic exposure control. Patients in all age groups were, on median, 25 to 35 mm below isocenter.

The pattern is the same in both: low patients get noisier posterior anatomy, and the dose rises.

How centering changes tube current modulation

Tube current modulation (GE AutomA and SmartmA, Siemens CARE Dose4D, and similar features on other systems) sets the mA from the patient's size and attenuation, which it reads from the scout.

The scout magnification problem

An AP or PA scout is a projection. A patient who sits closer to the tube looks larger on it, and a patient further from the tube looks smaller. The AAPM Alliance for Quality CT education slides on automatic exposure control explain that this is why miscentering affects AEC calculations for AP/PA scouts, and that lateral scouts are relatively immune from vertical miscentering.

So the system may plan the mA for the wrong-size patient. GE's dose guide makes the same point: off-centering distorts the size AutomA reads from the scout, so the mA is set from a less accurate estimate.

Which scout drives the mA

The same AAPM slides make three practical points for GE systems:

• If you take one scout, a lateral is the more robust choice for centering errors.
• If you take two, the most recent one drives the CTDIvol for the scan. The slides suggest AP/PA first for head and neck, and lateral first for chest, abdomen and pelvis.
• If you fix the table height after scouting, take new scouts. Your scout order and direction are part of your protocol, so check your own system's manual and your physicist's guidance.

A centering checklist for the scanner

Use this for every patient, not just the hard ones:

• Lasers on, and the patient's midline matched to the side laser at mid-thickness, not at the table top.
• Table height set so the body's center (front to back) sits at isocenter. Large patients are often set too low.
• Patient centered left to right on the table, with no rotation.
• Arms out of the scan field for body exams when the patient can tolerate it.
• Scout taken after the final position, in the direction your protocol calls for.
• Scout checked for centering before you plan the scan. If you move the table, re-scout.
• CTDIvol and DLP reviewed before scanning. If they look high for this patient, check centering first.

Here's how common centering errors show up:

GE's dose guide also notes that children are easier to place off-center than adults, and that good centering is what lets the filters and automatic exposure control work as designed. Take extra care with small patients.

Some newer scanners add centering aids; the AAPM slides mention a Patient Centering feature on several GE Revolution systems that gives feedback while you adjust the cradle. Aids help, but you still confirm the position and the scout.

Getting it consistent across your team

Centering is a habit, and habits drift. A few things help:

• Teach centering with the scout: show new staff what a low patient looks like and how the mA changes.
• Add a line to your protocol notes about which scout to take first.
• Review dose reports for outliers and check centering as part of the review.

If your team is new to a scanner, or you want a second set of eyes on scout and modulation settings, you can request applications training. For more on how reconstruction affects noise once the data is acquired, see our guide to ASiR-V and deep learning reconstruction on GE CT.

Any change to scout order, modulation settings or noise targets should go through your radiologist and medical physicist, and follow your facility's protocols.

Key takeaways

• The bowtie filter and tube current modulation both assume the patient is at isocenter.
• Most miscentered patients sit too low, which raises posterior noise and can raise dose.
• AP/PA scouts are affected by vertical miscentering; lateral scouts much less.
• Re-scout after you change table height.
• Check your lasers, and check every scout for centering before you plan.

FAQ

Why does CT patient centering affect radiation dose?

The bowtie filter and tube current modulation are designed for a patient at isocenter. Off-center patients can get the wrong mA from a distorted scout, and parts of the body get more or less beam than intended.

Is it worse to be too high or too low in the gantry?

Both cause problems. Studies show most patients are positioned too low, which increases noise in posterior anatomy and, with an AP/PA scout, can increase dose.

Should I use an AP or lateral scout for tube current modulation?

Lateral scouts are much less affected by vertical miscentering. Many protocols use both, and on some systems the most recent scout drives the mA, so the order matters. Follow your protocol and your physicist's guidance.

Do I need a new scout after I re-center the patient?

Yes. The mA plan comes from the scout, so if you move the table after scouting, take a new scout so the plan matches the patient's actual position.

How can I tell if a patient is off-center?

Check the lateral laser against the patient's mid-thickness, then check the scout. Some scanners also show a centering indicator or offer a camera-based aid.

Have a question about centering on your scanner? Reply or reach out, or get monthly support for help at the console when you need it.

Sources

• Kim MS, Singh S, Halpern E, Saini S, Kalra MK. Patient centering in abdominal CT, World Journal of Radiology, 2012
• Diagnostic Imaging (2014): Positioning essential for CT dose and image noise, summarizing Kaasalainen et al., American Journal of Roentgenology
• AAPM Alliance for Quality CT: Computed Tomography Automatic Exposure Control Education Slides (GE section)
• Manufacturer documentation: GE Healthcare, Blueprint: A Guide to CT Radiation Dose Management (2012)
• Manufacturer operator documentation: Siemens SOMATOM Perspective, CT Examination Instructions for Use, syngo CT VC40