Related Experiment Video
Updated: Jun 29, 2025
10:26
Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
Published on: May 19, 2023
1.8K
Graphical user interface-based convolutional neural network models for detecting nasopalatine duct cysts using
Kotaro Ito1, Naohisa Hirahara2, Hirotaka Muraoka2
1Department of Radiology, Nihon University School of Dentistry at Matsudo, 2-870-1 Sakaecho-Nishi, Matsudo, Chiba, 271-8587, Japan. itou.koutarou@nihon-u.ac.jp.
Scientific Reports
|April 2, 2024
Summary
A new deep learning model can detect nasopalatine duct cysts on panoramic radiographs, improving early diagnosis. This AI tool helps prevent these cysts from being overlooked in dental imaging.
Area of Science:
- Dentistry
- Medical Imaging
- Artificial Intelligence
Background:
- Nasopalatine duct cysts are often missed on panoramic radiographs due to image obstructions.
- Accurate detection of these cysts is crucial for timely clinical intervention.
Purpose of the Study:
- To develop a trained model for detecting nasopalatine duct cysts using panoramic radiography.
- To implement this model within a user-friendly graphical interface for clinical application.
Main Methods:
- A dataset of panoramic radiographs and CT images from 115 patients with nasopalatine duct cysts and 230 controls was used.
- Deep learning models (pretrained-LeNet and VGG16) were trained on 345 pre-processed panoramic radiographs.
- Model performance was evaluated based on accuracy, sensitivity, and specificity.
Main Results:
- The LeNet model achieved an accuracy of 85.3%.
- The VGG16 model demonstrated an accuracy of 88.0%.
- Both models showed potential for reliable cyst detection.
Conclusions:
- A simple deep learning approach can effectively train a model to detect nasopalatine duct cysts from panoramic radiographs.
- This AI-powered tool can assist clinicians in preventing overlooked diagnoses of these cysts.
- The developed system offers a practical solution for enhancing diagnostic sensitivity in dental imaging.
