# Impact of cranial collimation on cephalometric landmark visibility

## TL;DR

This study examines how cranial collimation affects the visibility of key landmarks in cephalometric analyses and identifies optimal settings to reduce radiation exposure without compromising diagnostic accuracy.

## Contribution

The study quantifies landmark masking frequency and derives ideal collimation parameters for diagnostic reliability.

## Key findings

- Nasion was masked in 0.6% of radiographs, while Sella remained visible in all.
- Limiting collimation to the upper 16.02% of image height could prevent most landmark masking.
- Head positioning and vertical skeletal growth patterns significantly influence landmark visibility.

## Abstract

To reduce radiation exposure, collimation of lateral cephalograms is frequently provided by manufacturers. However, cranial collimation geometries can lead to masking of anatomical reference points that are necessary for cephalometric analyses. To avoid unintended reference point masking, identifying optimal collimation geometries is crucial for diagnostic reliability and is therefore of significant clinical relevance. The aim of this study was to quantify (i) the frequency of Nasion and Sella landmark masking for a common collimation geometry and (ii) to derive ideal collimation parameters.

N = 1000 randomly selected lateral cephalograms were categorized into collimated (n = 49) and non-collimated (n = 951) images. The non-collimated images were virtually collimated using the same geometric settings. For each landmark and cephalogram, the frequency of landmark masking was assessed. For the statistical analysis log-regression was used to analyze association with head positioning, i.e. angular discrepancy to Frankfort horizontal plane (FHP-discrepancy), and Bjork sum growth pattern. The analyses were performed using the open-source software R using a significance level alpha of 0.05.

Sella was visible on all radiographs, whereas Nasion was masked in 0.6% of the lateral cephalograms. Masking could have been avoided if collimation was limited to the upper 16.02% of image height. Counter clockwise inclination of the head measured as FHP-discrepancy was significantly associated with landmark masking (p = 0.040) as well as a vertical skeletal growth pattern (as determined by the Bjork sum) (p = 0.029).

The evaluated collimation geometry that masks the upper 18% of the radiographs would have prevented masking for 99% of the included radiographs. Correct head positioning might further reduce the risk of landmark masking. The benefit of cranial collimation should be considered carefully for patients with pronounced vertical growth pattern (as determined by the Bjork sum).

The online version contains supplementary material available at 10.1186/s13005-026-00586-1.