Endoscopic System Based on Intraoral Camera and Image Processing

Objective:
In dentistry, dentists perform many treatments with a few visually magnified information. About deep area of root canals, it commonly relies on exploration carried out with the fingers of the dentist, because the root canal entrance is very small to be observed. In clinical practice, the intraoral camera and endoscopic systems are separate devices, and there are no sensors to capture both pictures of the whole tooth and those inside the root canal. Objective of this research is to combine the intraoral camera and endoscopic system facilitating to use clinical practice.
Methods:
We propose an endoscopic system based on the thin image fiber, SOPROLIFE intraoral camera as an image sensor, and a new adaptor to connect the two parts. We observed resolution charts with 50, 25, 10, and 5 line pairs (LP)/mm patterns. The acquired images are processed to both remove fixed-pattern noise using robust principal component analysis and enhance contrast.
Results:
The acquired images containing all LP/mm patterns were clear and showed higher contrast after processing. Visibility of the processed images is 1.7, 1.6, 2.2, and 1.9 times higher than that of the raw images for 50, 25, 10, and 5 LP/mm patterns, respectively.
Conclusion:
Our fabricated endoscopic system based on the SOPROLIFE intraoral camera could observe 50, 25, 10, and 5 LP/mm patterns. After image processing, the noise was reduced, and high-contrast images were obtained.
Significance:
This system can be considered as a further step toward facilitating noninvasive and contactless systems in clinical practice.