• Korean Journal of Radiology
• /
• v.22 no.3
• /
• pp.476-488
• /
• 2021

Objective: We aimed to develop a deep neural network for segmenting lung parenchyma with extensive pathological conditions on non-contrast chest computed tomography (CT) images. Materials and Methods: Thin-section non-contrast chest CT images from 203 patients (115 males, 88 females; age range, 31-89 years) between January 2017 and May 2017 were included in the study, of which 150 cases had extensive lung parenchymal disease involving more than 40% of the parenchymal area. Parenchymal diseases included interstitial lung disease (ILD), emphysema, nontuberculous mycobacterial lung disease, tuberculous destroyed lung, pneumonia, lung cancer, and other diseases. Five experienced radiologists manually drew the margin of the lungs, slice by slice, on CT images. The dataset used to develop the network consisted of 157 cases for training, 20 cases for development, and 26 cases for internal validation. Two-dimensional (2D) U-Net and three-dimensional (3D) U-Net models were used for the task. The network was trained to segment the lung parenchyma as a whole and segment the right and left lung separately. The University Hospitals of Geneva ILD dataset, which contained high-resolution CT images of ILD, was used for external validation. Results: The Dice similarity coefficients for internal validation were 99.6 ± 0.3% (2D U-Net whole lung model), 99.5 ± 0.3% (2D U-Net separate lung model), 99.4 ± 0.5% (3D U-Net whole lung model), and 99.4 ± 0.5% (3D U-Net separate lung model). The Dice similarity coefficients for the external validation dataset were 98.4 ± 1.0% (2D U-Net whole lung model) and 98.4 ± 1.0% (2D U-Net separate lung model). In 31 cases, where the extent of ILD was larger than 75% of the lung parenchymal area, the Dice similarity coefficients were 97.9 ± 1.3% (2D U-Net whole lung model) and 98.0 ± 1.2% (2D U-Net separate lung model). Conclusion: The deep neural network achieved excellent performance in automatically delineating the boundaries of lung parenchyma with extensive pathological conditions on non-contrast chest CT images.

• The Journal of the Korea Contents Association
• /
• v.16 no.2
• /
• pp.558-565
• /
• 2016

Recent 3D printing technology has been grafting onto various medical practices. In light of this trend, this research is intended to examine the figuration surface's accuracy of 3D images made by using DICOM images after printing by 3D printing. The medical images were obtained from animal bone objects, while the objects were printed after undergoing STL file conversion for 3D printing purposes. Ultimately, after the 3D figuration, which was obtained by the original animal bones and 3D printing, was scanned by 3D scanner, 3D modeling was merged each other and the differences were compared. The result analysis was conducted by visual figuration comparison, color comparison of modeling's scale value, and numerical figuration comparison. The shape surface was not visually distinguished; the numerical figuration comparison was made from the values measured from the four different points on the X, Y and Z coordinates. The shape surface of the merged modeling was smaller than the original object (the animal bone) by average of -0.49 mm in the 3D printed figuration. However, not all of the shape surface was uniformly reduced in size and the differences was within range of -0.83 mm on the experiment.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.5
• /
• pp.991-996
• /
• 2007

Recently, a computational reconstruction method using an integral imaging technique, which is a promise three-dimensional display technique, has been actively researched. This method is that 3-D images can be digitally reconstructed at the required output planes by superposition of all of the inversely enlarged elemental images by using a hypothetical pinhole array model. However, the conventional method mostly yields reconstructed images having a low-resolution, because there are some intensity irregularities with a grid structure at the reconstructed mage plane by using square-type elemental images. In this paper, to overcome this problem, we propose a novel computational integral imaging reconstruction (CIIR) method using round-type mapping model. Proposed CIIR method can overcome problems of non-uniformly reconstructed images caused from the conventional method and improve the resolution of 3-D images. To show the usefulness of the proposed method, both computational experiment and optical experiment are carried out and their results are presented.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.6
• /
• pp.621-629
• /
• 2011

According to the development of technologies for generating the 3D spatial information, the needs for producing and updating the precise 3D objects with LoD 4 level are increased. On the other hand, the needs for real-time updating of 2D digital maps are expanded, based on the execution of various GIS projects. These 2D informations can be extracted from precisely constructed 3D spatial information, to do this the feasibility studies on extraction of the 2D information from the 3D spatial information is needed. In this study, 3D objects are modeled using multi-oblique images, and the objects are stereo-plotted using digital airborne images, as well. Then the two data sets are compared and analyzed. The results show that the accuracy assessments fulfill the 1/1,000 digital map accuracy standard of regulations for photogrametric surveying of National Geographic Information Institute, but the shapes and the areas of building objects are different between two data sets because of the portrayal standards. Consequently, researchers can conclude that it is possible to generate the building layer of large scale topographic map using multi-oblique images, but additional researches is needed to resolve the problems on differences of the portrayal standards.

• Imaging Science in Dentistry
• /
• v.33 no.3
• /
• pp.151-159
• /
• 2003

Purpose: The purpose of this study is to evaluate the effectiveness and usefulness of newly developed personal computer-based software to eliminate the linear artifacts by the metal restorations. Materials and Methods: A 3D CT image was conventionally reconstructed using ADVANTAGE WINDOWS 2.0 3D Analysis software (GE Medical System, Milwaukee, USA) and eliminated the linear artifacts manually. Next, a 3D CT image was reconstructed using V-works 4.0/sup TM/(Cybermed Inc., Seoul, Korea) and the linear artifacts eliminated manually in the axial images by a skillful operator using a personal computer. A 3D CT image was reconstructed using V-works 4.0/sup TM/(Cybermed Inc., Seoul, Korea) and the linear artifacts were removed using a simplified algorithm program to eliminate the linear artifacts automatically in the axial images using a personal computer, abbreviating the manual editing procedure. Finally, the automatically edited reconstructed 3D images were compared to the manually edited images. Results and Conclusion: We effectively eliminated the linear artifacts automatically by this algorithm, not by the manual editing procedures, in some degree. But programs based on more complicated and accurate algorithms may lead to a nearly flawless elimination of these linear artifacts automatically.

• Journal of Multimedia Information System
• /
• v.9 no.3
• /
• pp.183-190
• /
• 2022

Image-based 3D object detection is one of the important and difficult problems in autonomous driving and robotics, and aims to find and represent the location, dimension and orientation of the object of interest. It generates three dimensional (3D) bounding boxes with only 2D images obtained from cameras, so there is no need for devices that provide accurate depth information such as LiDAR or Radar. Image-based methods can be divided into three main categories: monocular, stereo, and multi-view 3D object detection. In this paper, we investigate the recent state-of-the-art models of the above three categories. In the multi-view 3D object detection, which appeared together with the release of the new benchmark datasets, NuScenes and Waymo, we discuss the differences from the existing monocular and stereo methods. Also, we analyze their performance and discuss the advantages and disadvantages of them. Finally, we conclude the remaining challenges and a future direction in this field.

• Journal of Internet Computing and Services
• /
• v.10 no.4
• /
• pp.25-33
• /
• 2009

In the paper, we proposed a new system to retrieve 3D wrinkle data based on stereo images. Usually, 3D reconstruction based on stereo images or video is very popular and it is the research focus, which has been applied for culture heritage, building and other scene. The target is object measurement, the scene depth calculation and 3D data obtained. There are several challenges in our research. First, it is hard to take the full information wrinkle images by cameras because of light influence, skin with non-rigid object and camera performance. We design a particular computer vision system to take winkle images with a long length camera lens. Second, it is difficult to get the dense stereo data because of the hard skin texture image segmentation and corner detection. We focus on semi-dense stereo matching algorithm for the wrinkle depth. Compared with the 3D scanner, our system is much cheaper and compared with the physical modeling based method, our system is more flexible with high performance.

• The Journal of the Convergence on Culture Technology
• /
• v.6 no.4
• /
• pp.803-806
• /
• 2020

The body shape measurement method using 3D images has been widely used due to the recent development of 3D measurement cameras and algorithms. Existing 3D imaging devices are expensive devices, and there is a limit to their universalization. Due to the recent spread of inexpensive 3D cameras and the development of various measurement methods, various possibilities are being shown. It is expected to have a great impact on the medical device market that requires accurate data collection. Various medical device products using artificial intelligence are emerging, and accurate data collection is the most important to develop accurate artificial intelligence algorithms. Collection equipment using 3D cameras is expected to act as a major factor in the development of artificial intelligence algorithms using 3D images.

• Korean Journal of Remote Sensing
• /
• v.28 no.3
• /
• pp.319-327
• /
• 2012

In this paper, we introduce a newly developed photogrammetric software for automatic generation of high quality geo-products and its performance assessment carried out using various satellite images. Our newly developed software provides the latest techniques of an optimized sensor modelling, ortho-image generation and automated Digital Elevation Model (DEM) generation for diverse remote sensing images. In particular, images from dual- and multi-sensor images can be integrated for 3D mapping. This can be a novel innovation toward a wider applicability of remote sensing data, since 3D mapping has been limited within only single-sensor so far. We used Kompsat-2, Ikonos, QuickBird, Spot-5 high resolution satellite images to test an accuracy of 3D points and ortho-image generated by the software. Outputs were assessed by comparing reliable reference data. From various sensor combinations 3D mapping were implemented and their accuracy was evaluated using independent check points. Model accuracy of 1~2 pixels or better was achieved regardless of sensor combination type. The high resolution ortho-image results are consistent with the reference map on a scale of 1:5,000 after being rectified by the software and an accuracy of 1~2 pixels could be achieved through quantitative assessment. The developed software offers efficient critical geo-processing modules of various remote sensing images and it is expected that the software can be widely used to meet the demand on the high-quality geo products.

• Korean Journal of Computational Design and Engineering
• /
• v.16 no.3
• /
• pp.216-226
• /
• 2011

Surgeon dentists usually rely on their experiential judgments from patients' oral plaster casts and medical images to determine the positional and directional information of implant fixtures and to perform drilling tasks during dental implant surgical operations. This approach, however, may cause some errors and deteriorate the quality of dental implants. Computer-aided methods have been introduced as supportive tools to alleviate the shortcomings of the conventional approach. In this paper, we present an approach of 3D dental implant simulation which can provide the realistic and immersive experience of dental implant information. The dental implant information is primarily composed of several kinds of 3D mesh models obtained as follows. Firstly, we construct 3D mesh models of jawbones, teeth and nerve curves from the patient's dental images using software $Mimics^{TM}$. Secondly, we construct 3D mesh models of gingival regions from the patient's oral impression using a reverse engineering technique. Thirdly, we select suitable types of implant fixtures from fixture database and determine the positions and directions of the fixtures by using the 3D mesh models and the dental images with software $Simplant^{TM}$. Fourthly, from the geometric and/or directional information of the jawbones, the gingival regions, the teeth and the fixtures, we construct the 3D models of surgical guide stents which are crucial to perform the drilling operations with ease and accuracy. In the application phase, the dental implant information is combined with the tangible interface device to accomplish 3D dental implant simulation. The user can see and touch the 3D models related with dental implant surgery. Furthermore, the user can experience drilling paths to make holes where fixtures are implanted. A preliminary user study shows that the presented approach can be used to provide dental students with good educational contents. With future work, we expect that it can be utilized for clinical studies of dental implant surgery.