# Effect of CBCT voxel size and software program on the detection of fenestrations and dehiscences

## TL;DR

This study found that CBCT imaging accuracy for detecting dental fenestrations and dehiscences is not significantly affected by voxel size or software used, but performance is better for fenestrations than dehiscences.

## Contribution

The study demonstrates that diagnostic accuracy for fenestrations and dehiscences is not significantly influenced by CBCT voxel size or software program.

## Key findings

- CBCT showed high accuracy (94%) for detecting fenestrations with low false positives.
- Dehiscence detection had lower accuracy (70%) and higher false positives (33–40%).
- Voxel size and software had little impact on diagnostic performance.

## Abstract

This study investigated the effect of CBCT voxel size and software program on the accuracy of detecting fenestrations and dehiscences.

This ex-vivo experimental animal study was carried out on 6 fleshy sheep heads with both maxilla and mandible accompanied with the surrounding soft tissue. Fenestrations [n = 48] and dehiscences [n = 51] were randomly created on the buccal surfaces of teeth. Two CBCT scans were acquired for each sheep head using two different voxel sizes [0.2 mm and 0.4 mm]. Images were analyzed by two oral radiologists on two different software programs [OnDemand and ImplaStation]. All data was collected and analyzed in terms of diagnostic performance measures.

For fenestrations, both voxel sizes and software programs showed high diagnostic performance, with approximately 94% accuracy and sensitivity, 95% specificity, and a low false positive rate of about 5%. For dehiscences, performance was more limited, with around 70% accuracy and sensitivity, 60–67% specificity, and a higher false positive rate of 33–40%. None of the investigated parameters showed statistically significant differences.

Changes in voxel size and software program had little impact on the diagnostic accuracy for detecting fenestrations and dehiscences. CBCT demonstrated good performance in identifying fenestrations but only moderate capability in detecting dehiscences. These results indicate that, although CBCT is a useful imaging tool, its contribution to diagnosing dehiscences is limited by factors such as volume averaging and partial volume effects, and should therefore be interpreted with caution in clinical settings.

Clinically, the findings suggest that choosing between 0.2 mm and 0.4 mm voxel sizes, or between different viewing software programs OnDemand and ImplaStation, does not meaningfully affect the ability to diagnose fenestrations and dehiscences. This provides flexibility in routine practice, allowing clinicians to select imaging settings based on practical considerations such as field of view, dose, or workflow, without compromising diagnostic confidence.