Key takeaways
- Record dimensions, units and the scan mode together.
- Change one grid input at a time.
- Keep sample-size arithmetic separate from clinical review.
What does CBCT mean?
The FDA describes a rotating cone-shaped X-ray beam whose data reconstruct three-dimensional anatomy. Its dental CBCT guidance says use should be limited to necessary clinical information that other imaging methods cannot provide. A purchase worksheet cannot decide whether a patient's scan is justified.
Separate the volume from its grid
Ask the seller or vendor to identify the exact model, scan mode, stated field of view and reconstructed grid. Do not treat a device name as proof of every advertised mode. Use the equipment value guide for the asset and condition review.
Duke's MorphoSource documentation explains that stacked image slices form a volume, with spacing measured along three axes. A voxel is a volume element in that grid. Record the units with each dimension; a count of samples and a physical length answer different questions.
The DICOM encoding standard distinguishes bits allocated from bits stored. Its example allocates 16 bits, or two bytes, to a sample even when fewer bits carry its value. Do not substitute one field for the other in a data-size worksheet.
A deliberately simplified grid example
Assume an invented cubic reconstructed volume with cells that exactly fill each side. Each cell holds one uncompressed sample using two bytes. Spacing is the same in all three directions. These are mathematical inputs, not specifications for an actual CBCT unit or a recommended scanning protocol.
| Independent grid case | Samples on each axis | Total samples | Sample bytes only |
|---|---|---|---|
| 80 mm per side; 0.2 mm cells | 400 | 64,000,000 | 128,000,000 |
| Same cube; only cell width changes to 0.4 mm | 200 | 8,000,000 | 16,000,000 |
| Original cells; only cube side doubles to 160 mm | 800 | 512,000,000 | 1,024,000,000 |
For the first row, divide 80 by 0.2, multiply 400 by itself three times, then multiply by two bytes. Doubling the cell width divides this count by eight. Doubling each side at the original width multiplies it by eight. Each sensitivity starts from the first row.
The worksheet excludes headers, compression, source projections, extra series and backups. It is not an actual DICOM file size or a storage budget. A real scan volume need not be cubic; obtain its actual dimensions and sample counts before reusing the formula.
Common mistakes
Do not use the smallest displayed cell size as a verdict on clinical image quality. Keep clinical review separate from this arithmetic. The FDA's guidance ties technique choices to the clinical question, patient size and scanned area, with proper equipment maintenance and testing.
Use buyer diligence to assign the technical review. Keep any proposed upgrade in the equipment financing plan, and assess its separate cash need.
Summary
Keep the 80 mm dimension, 0.2 mm cell width and two-byte assumption together. Changing one input changes the sample count in a predictable way. Bring the actual specifications and unresolved review items to a purchase discussion, without treating the worksheet as clinical approval.
Frequently asked questions
Why does doubling cell width reduce samples eightfold?
In the invented cube, each of three axis counts halves. One-half multiplied three times is one-eighth.
Is 128,000,000 bytes an actual scan file size?
No. It counts only the assumed sample bytes and excludes headers, compression and other data.
Does the 160 mm row change cell width too?
No. It independently doubles each side from the base case while retaining 0.2 mm cells.
Does this worksheet establish an equipment price?
No. It explains grid inputs. Ownership, condition, costs and the supported purchase terms need their own review.
Sources
Retrieval dates appear beside each source. Figures retain their stated observation years; retrieval does not make older data current.
- FDA dental CBCT guidance · Retrieved
- Duke volume spacing · Retrieved
- DICOM pixel encoding · Retrieved