How to Measure Your CBCT Image Quality: A Technical Checklist for Your Team

Why You Should Audit Your Own Image Quality

Most imaging centers trust that their scanner "works fine" because it produces images that look okay. That's not enough. Image quality degrades gradually: detector drift, geometry miscalibration, collimator dust, tube aging. By the time a dentist complains, it's already too late — months of suboptimal studies are behind you.

A monthly 45-minute internal audit can catch degradation before your referring dentists do. Here's the checklist.

1. Voxel Size

The voxel is the 3D pixel size, expressed in microns (μm). For dental imaging, typical ranges are:

• Implants and endodontics: 150 to 200 μm.
• Complex endodontics (apical microsurgery, fractures): 75 to 125 μm.
• Orthodontics and general maxillofacial: 200 to 400 μm.

Check monthly that each case type's protocol is using the correct voxel size. It's common to find centers using 300 μm for everything, including endodontics, simply because "that's the default preset."

2. FOV (Field of View)

The FOV defines which anatomical region you capture. Bigger isn't better — it increases patient dose and reduces useful resolution per voxel.

• Small FOV (5x5 cm): localized area, single implant, endodontics.
• Medium FOV (8x8 to 10x10 cm): partial arch, third molars, multiple implant planning.
• Large FOV (15x15 to 20x17 cm): full maxillofacial, TMJ, orthodontics with 3D cephalometric analysis.

The ALARA principle (As Low As Reasonably Achievable) requires using the minimum FOV sufficient for the clinical question. Document the justification in every protocol.

3. Contrast and Noise

An image with low contrast or high noise makes it hard to differentiate structures. To audit:

• Use a standard phantom (SedentexCT-type models or equivalent, if budget allows). If not, an improvised phantom made from PMMA cylinders of known densities works for tracking trends.
• Measure contrast-to-noise ratio (CNR) monthly on the same phantom with the same protocol.
• Log the values in a spreadsheet. If they drop more than 15% from the historical average, call technical support.

4. Artifacts

The most common artifacts in dental CBCT are:

• Beam hardening: dark bands between metallic structures (implants, crowns, amalgam restorations).
• Patient motion: bent edges, visible blurring in sagittal slices.
• Ring artifacts: concentric circles caused by defective detector pixels.
• Truncation: when anatomy extends beyond the FOV and the reconstruction shows dark edges.

Of these, ring artifacts are the clearest warning sign of a technical problem: if they appear, calibrate the detector immediately.

5. Geometry and Measurement Accuracy

Linear measurements need to be exact. This is critical for implant planning (measuring available bone height to determine implant length) and for orthodontics. A good monthly test:

1. Scan a phantom with objects of known dimensions (e.g., a metal ruler or certified-size spheres).
2. Measure in the viewer across all three planes.
3. Error should be under 1%. Above 2%, recalibrate geometry.

6. Dose (CTDIvol and DAP)

Dose isn't strictly an image quality parameter, but it is a service quality one. Verify that your protocols stay within typical ranges:

• Small-FOV dental CBCT: 40 to 90 μSv.
• Large-FOV CBCT: 80 to 250 μSv.

If your values run higher, review kVp, mAs, number of projections, and exposure time. Lowering dose without losing diagnostic quality is one of the most important ethical differentiators you can offer.

7. Consistency Across Technologists

Audit monthly: pull 10 random studies from the month, separated by technologist. Are they using the same protocols? Was patient positioning consistent? Differences between technologists are usually the biggest source of quality variability.

The Complete Monthly Protocol (45 Minutes)

1. Phantom scan with standard protocol (10 min).
2. CNR, noise, and artifact measurement on the phantom (10 min).
3. Review of 10 real studies from the month (15 min).
4. Log results and compare to historical data (10 min).

When to Escalate to Technical Support

• CNR drop of more than 15% across two consecutive measurements.
• Ring artifacts appear.
• Geometric error above 2%.
• Any repeated complaint from referring dentists about sharpness.

Once your technical quality is solid, the next step is making sure the dentist can actually appreciate it. If you deliver excellent images but the viewer is poor, it doesn't show. You can try CBCTHub free and see how your best studies look in a modern viewer.