# Artificial intelligence in diagnosis of maxillary sinusitis: A clinical study

## TL;DR

This study shows that artificial intelligence is highly reliable in diagnosing maxillary sinusitis using CBCT scans, improving diagnostic accuracy and efficiency.

## Contribution

The study demonstrates AI's high diagnostic performance for maxillary sinusitis using CBCT scans in a clinical setting.

## Key findings

- AI achieved 89.1% sensitivity and 91.7% specificity in diagnosing maxillary sinusitis.
- The model had an overall accuracy of 90.4% and an area under the ROC curve of 0.948.
- AI shows potential as an effective tool to enhance diagnostic consistency in clinical imaging.

## Abstract

Maxillary sinusitis is a common inflammatory condition often diagnosed through imaging modalities such as "cone-beam computed
tomography (CBCT)". A prospective clinical research was conducted involving 200 patients with suspected maxillary sinusitis. The AI
model achieved a sensitivity of 89.1%, specificity of 91.7%, positive predictive value of 92.5% and negative predictive value of 87.8%
and overall accuracy of 90.4%. The area under the ROC curve was 0.948, indicating excellent diagnostic discrimination. AI demonstrates
high reliability in diagnosing maxillary sinusitis on CBCT scans and may serve as an effective adjunct in clinical imaging interpretation,
enhancing diagnostic efficiency and consistency.