• Imaging Science in Dentistry
• /
• v.52 no.4
• /
• pp.383-391
• /
• 2022

Purpose: Periodontitis, the most prevalent chronic inflammatory condition affecting teeth-supporting tissues, is diagnosed and classified through clinical and radiographic examinations. The staging of periodontitis using panoramic radiographs provides information for designing computer-assisted diagnostic systems. Performing image segmentation in periodontitis is required for image processing in diagnostic applications. This study evaluated image segmentation for periodontitis staging based on deep learning approaches. Materials and Methods: Multi-Label U-Net and Mask R-CNN models were compared for image segmentation to detect periodontitis using 100 digital panoramic radiographs. Normal conditions and 4 stages of periodontitis were annotated on these panoramic radiographs. A total of 1100 original and augmented images were then randomly divided into a training (75%) dataset to produce segmentation models and a testing (25%) dataset to determine the evaluation metrics of the segmentation models. Results: The performance of the segmentation models against the radiographic diagnosis of periodontitis conducted by a dentist was described by evaluation metrics(i.e., dice coefficient and intersection-over-union [IoU] score). MultiLabel U-Net achieved a dice coefficient of 0.96 and an IoU score of 0.97. Meanwhile, Mask R-CNN attained a dice coefficient of 0.87 and an IoU score of 0.74. U-Net showed the characteristic of semantic segmentation, and Mask R-CNN performed instance segmentation with accuracy, precision, recall, and F1-score values of 95%, 85.6%, 88.2%, and 86.6%, respectively. Conclusion: Multi-Label U-Net produced superior image segmentation to that of Mask R-CNN. The authors recommend integrating it with other techniques to develop hybrid models for automatic periodontitis detection.

• Journal of Korean Academy of Oral and Maxillofacial Radiology
• /
• v.16 no.1
• /
• pp.59-68
• /
• 1986

The purpose of this investigation was to locate the central plane of the image layer on the panoramic machine relative to a specific point on the machine. In the study of the central plane of the image layer of panoramic radiograph, using the Morrita Company PANEX-EC a series of 33 exposures were taken with the 4-5 experimental pins placed in the holes of the plastic model plate, then evaluated by human eye. The author analyzed the central plane of the image layer by Mitutoyo-A-221 and calculated horizontal and vertical magnification ratio in the central plane of the image layer determined experimentally. The results were as follows: 1. The location of the central plane of the image layer determined experimentally was to lateral, compared with manufactural central plane. 2. Horizontal magnification ratio in the central plane of the image layer determined experimentally was 9.25%. 3. Vertical magnification ratio in the central plane of the image layer determined experimentally was 9.17%.

• Imaging Science in Dentistry
• /
• v.33 no.3
• /
• pp.131-135
• /
• 2003

Purpose: To determine whether the mandibular radiomorphometric indices in panoramic radiography are correlated with the bone mineral density of Cu-equivalent images in intraoral film. Materials and Methods: The bone mineral density (BMD) of the mandibular premolar area was measured in the Cu-equivalent image of intraoral film. The Panoramic Mandibular Index (PMI) and Mandibular Cortical Width (MCW) were measured in panoramic radiographs of six dry mandibles, and the Pearson correlation between PMI, MCW, and BMD were tested. Results: There were no significant correlations between PMI and BMD (r = 0.280), nor between MCW and BMD (r =0.237). Conclusion: The results show that PMI and MCW were poor diagnostic indicators of mandibular BMD in the six dry mandibles used in this study. The correlationship between the mandibular radiomorphometric indices (PMI and MCW) and mandibular BMD needs to be researched further using large in vivo patient samples.

• Imaging Science in Dentistry
• /
• v.33 no.4
• /
• pp.199-205
• /
• 2003

Purpose: To compare the visibility of the mandibular canal at the different radiographic methods such as conventional panoramic radiographs, Vimplant multi planar reformatting (MPR)-CT panoramic images, Vimplant MPR-CT paraxial images and film-based DentaScan MPR-CT images. Materials and Methods: Data of 11 mandibular dental implant patients, who had been planned treatment utilizing both panoramic and MPR-CT examination with DentaScan software (GE Medical systems, Milwaukee, USA), were used in this study. The archived axial CT data stored on CD-R discs were transferred to a personal computer with 17' LCD monitor. Paraxial and panoramic images were reconstructed using Vimplant software (CyberMed Inc., Seoul, Korea). Conventional panoramic radiographs, monitor-based Vimplant MPR-CT panoramic images, monitor-based Vimplant MPR-CT paraxial images, and film-based DentaScan MPR-CT images were evaluated for visibility of the mandibular canal at the mental foramen, 1 cm, 2 cm, and 3 cm posterior to mental foramen using the 4-point grading score. Results: Vimplant MPR-CT panoramic, paraxial, and DentaScan MPR-CT images revealed significantly clearer images than conventional panoramic radiographs. Particularly at the region 1 em posterior to mental foramen, conventional panoramic radiographs showed a markedly lower percentage of 'excellent' mandibular canal images than images produced by other modalites. Vimplant MPR-CT and DentaScan MPR-CT images did not show significant difference in visibility of the mandibular canal. Conclusion: The study results show that Vimplant and DentaScan MPR-CT imaging systems offer significantly better images of the mandibular canal than conventional panoramic radiograph.

• Imaging Science in Dentistry
• /
• v.48 no.4
• /
• pp.255-259
• /
• 2018

Purpose: This study proposes a new ball-type phantom for evaluation of the image layer of panoramic radiography. Materials and Methods: The arch shape of an acrylic resin phantom was derived from average data on the lower dental arch in Korean adult males. Metal balls with a 2-mm diameter were placed along the center line of the phantom at a 4-mm mesiodistal interval. Additional metal balls were placed along the 22 arch-shaped lines that ran parallel to the center line at 2-mm buccolingual intervals. The height of each ball in the horizontal plane was spaced by 2.5 mm, and consequently, the balls appeared oblique when viewed from the side. The resulting phantom was named the Panorama phantom. The distortion rate of the balls in the acquired image was measured by automatically calculating the difference between the vertical and horizontal length using $MATLAB^{(R)}$. Image layer boundaries were obtained by applying various distortion rate thresholds. Results: Most areas containing metal balls (91.5%) were included in the image layer with a 50% distortion rate threshold. When a 5% distortion rate threshold was applied, the image layer was formed with a small buccolingual width along the arch-shaped center line. However, it was medially located in the temporomandibular joint region. Conclusion: The Panorama phantom could be used to evaluate the image layer of panoramic radiography, including all mesiodistal areas with large buccolingual width.

• Journal of KIISE
• /
• v.42 no.8
• /
• pp.990-997
• /
• 2015

Panoramic image creation has been extensively studied. Existing methods use customized hardware, or apply post-processing methods to seamlessly stitch images. These result in an increase in either cost or complexity. In addition, images can only be stitched under certain conditions such as existence of characteristic points of the images. This paper proposes a low cost and easy-to-use system that produces realtime panoramic video. We use an off-the-shelf embedded platform to capture multiple images, and these are then transmitted to a server in a compressed format to be merged into a single panoramic video. Finally, we analyze the performance of the implemented system by measuring time to successfully create the panoramic image.

• Imaging Science in Dentistry
• /
• v.38 no.4
• /
• pp.209-213
• /
• 2008

Purpose : To evaluate the relative diagnostic accuracy of panoramic radiography and Water's projection in maxillary sinus inflammatory diseases by comparing the radiodensities of the images with those of CT. Materials and Methods : Panoramic radiographs, Waters' projection, and CT images from 55 subjects (11O sinuses) were included in this retrospective study. The radiodensity of each maxillary sinus in panoramic radiography was recorded separately as upper and lower divided horizontally by hard palate. In Waters' projection, the overall sinus radiodensity was recorded. The CT images were considered as gold standard. Results : In panoramic radiography, 83 sinuses had same upper and lower radiodensity and 72 of these were consistent with those of CT, 26 sinuses had different upper and lower radiodensity and 15 of these, upper radiodensity was consistent with CT, the remaining 11, lower radiodensity was consistent with CT. One sinus had upper radiolucency with lower radiopacity and both were consistent with those of CT. Altogether 73 (66.4%) among 110 sinuses in panoramic radiography showed full agreement with CT, 26 (23.6%) showed partial agreement with CT. 9 sinuses had no lower image under the hard palate in panoramic radiography due to the smaller size of sinus. In Waters' projection, the radiodensity of 105 sinuses (95.5%) were consistent with that of CT. Conclusion : The panoramic radiography showed 90.0% of the sinus conditions fully or partially which may appear less accurate than that of Water's view (95.5%) but with more detailed information of the inferior part of sinuses. (Korean J Oral Maxillofac Radiol 2008; 38 : 209-13)

• Journal of Korean Academy of Oral and Maxillofacial Radiology
• /
• v.28 no.1
• /
• pp.47-58
• /
• 1998

The purpose of this study was to investigate an influence of the change of central panoramic curves on the image reconstruction in the dental implant CT. The author designed three experimental groups according to the location of central panoramic curve. In group A, central panoramic curve was determined as the curve connecting the center of roots from the first premolar to the first molar. In group B, central panoramic curve was determined as the line connecting the lingual cortical plate at the level of the mesial aspect of the first premolar with the buccal cortical plate at the level of the mesial aspect of the first molar. In Group C, central panoramic curve was determined as the line connecting the buccal cortical plate at the level of the mesial aspect of the first premolar with the lingual cortical plate at the level of the mesial aspect of the first molar. Twenty four reformatted CT images was acquired from four mandibles embedded in the resin block and twenty four contact radiographs of dog specimens were acquired. Each Image was processed under Adobe Photoshop program analysed by MSPA(mandible/maxilla shape pattern analysis) variables such as MXVD, MXHD, UHD, MHD, and LHD. The obtained results were as follows ; 1. The mean of MXVD variable was 19.9, 20.2, and 20.0 in group A, B, and C, respectively, which were smaller than actual value 20.5. But, there was no significant difference among 3 groups (p>0.05). 2. The mean of MXHD, UHD, MHD, and LHD variables in group A, B, and C was 11.9, 12.2, and 12.3; 9.3, 9.5, and 9.6; 10.0, 10.3, and 10.3; 9.2, 9.3, and 9.4 respectively which were equal to or greater than the actual value 11.8, 9.3, 10.0, and 9.2. But, there was no significant difference among 3 groups (p>0.05). 3. The number of noneffective observations with difference over or under 1 mm with comparison to the actual value was 24(20％), 58(48.3％), and 52(43.3％), respectively, in group A, B, and C. 4. In group A, the number of observations over 1 mm and under 1 mm was 9 and 15, respectively, but in group Band C, the number of observations over 1 mm was more than under 1 mm.

• The Journal of the Korea Contents Association
• /
• v.17 no.3
• /
• pp.42-52
• /
• 2017

Panorama image is a single image obtained by combining images taken at several viewpoints through matching of corresponding points. Existing panoramic image generation methods that find the corresponding points are extracting local invariant feature points in each image to create descriptors and using descriptor matching algorithm. In the case of video sequence, frames may be a lot, so therefore it may costs significant amount of time to generate a panoramic image by the existing method and it may has done unnecessary calculations. In this paper, we propose a method to quickly create a single panoramic image from a video sequence. By assuming that there is no significant changes between frames of the video such as in locally, we use the FAST algorithm that has good repeatability and high-speed calculation to extract feature points and the Lucas-Kanade algorithm as each feature point to track for find the corresponding points in surrounding neighborhood instead of existing descriptor matching algorithms. When homographies are calculated for all images, homography is changed around the center image of video sequence to warp images and obtain a planar panoramic image. Finally, the spherical panoramic image is obtained by performing inverse transformation of the spherical coordinate system. The proposed method was confirmed through the experiments generating panorama image efficiently and more faster than the existing methods.

• Journal of KIISE:Software and Applications
• /
• v.36 no.8
• /
• pp.653-663
• /
• 2009

In this paper, we propose a blind spots elimination system using three cameras. A wide-angle camera is attached on trunk for eliminating blind spots of a rear-view mirror and two cameras are attached on each side-view mirror for eliminating blind spots of vehicle's sides. In order to eliminate blind spots efficiently, we suggest a method to build a panoramic mosaic view with two side images and one wide-angle rear image. First, we obtain an undistorted image from a wide-angle camera of rear-view and calculate the focus-of-contraction (FOC) in undistorted images of rear-view while the car is moving straight forward. Second, we compute a homography among side-view images and an undistorted image of rear-view in flat road scenes. Next, we perform an image registration process after road and background region segmentation. Finally, we generate various views such as a cylinder panorama view, a top view and an information panoramic mosaic view.