• Journal of IKEEE
• /
• v.26 no.3
• /
• pp.408-415
• /
• 2022

In this paper, we propose a 3D point cloud reconstruction technique from 2D images using efficient feature map extraction network. The originality of the method proposed in this paper is as follows. First, we use a new feature map extraction network that is about 27% efficient than existing techniques in terms of memory. The proposed network does not reduce the size to the middle of the deep learning network, so important information required for 3D point cloud reconstruction is not lost. We solved the memory increase problem caused by the non-reduced image size by reducing the number of channels and by efficiently configuring the deep learning network to be shallow. Second, by preserving the high-resolution features of the 2D image, the accuracy can be further improved than that of the conventional technique. The feature map extracted from the non-reduced image contains more detailed information than the existing method, which can further improve the reconstruction accuracy of the 3D point cloud. Third, we use a divergence loss that does not require shooting information. The fact that not only the 2D image but also the shooting angle is required for learning, the dataset must contain detailed information and it is a disadvantage that makes it difficult to construct the dataset. In this paper, the accuracy of the reconstruction of the 3D point cloud can be increased by increasing the diversity of information through randomness without additional shooting information. In order to objectively evaluate the performance of the proposed method, using the ShapeNet dataset and using the same method as in the comparative papers, the CD value of the method proposed in this paper is 5.87, the EMD value is 5.81, and the FLOPs value is 2.9G. It was calculated. On the other hand, the lower the CD and EMD values, the better the accuracy of the reconstructed 3D point cloud approaches the original. In addition, the lower the number of FLOPs, the less memory is required for the deep learning network. Therefore, the CD, EMD, and FLOPs performance evaluation results of the proposed method showed about 27% improvement in memory and 6.3% in terms of accuracy compared to the methods in other papers, demonstrating objective performance.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.26 no.1
• /
• pp.27-32
• /
• 2022

Purpose When performing a whole-body bone scan, many patients are experiencing psychological difficulties due to the close distance to the detector. Recently, in the medical field, there is a report that using virtual reality (VR) equipment can give pain relief to pediatric patients with weak concentration or patients receiving severe treatment through a distraction method. Therefore, in this paper, VR equipment was used to provide psychological stability to patients during nuclear medicine tests, and it is intended to evaluate whether it can be used in clinical practice. Materials and Methods As VR equipment, ALLIP Z6 VR (ALLIP, Korea) was used and the experiment was conducted after connecting to a mobile phone. The subjects were 30 patients who underwent whole-body bone examination from September 1, 2021 to September 30, 2021. After intravenous injection of 99mTc-HDP, 3 to 6 hours later, VR equipment was put on and whole body images were obtained. After the test, a survey was conducted, and a Likert scale of 5 points was used for psychological anxiety and satisfaction with VR equipment. Hypothesis verification and reliability of the survey were analyzed using SPSS Statistics 25 (IBM, Corp., Armonk, NY, USA). Results Anxiety about the existing whole-body bone test was 3.03±1.53, whereas that of anxiety after wearing VR equipment was 2.0±1.21, indicating that anxiety decreased to 34%. When regression analysis of the effect of the patient's concentration on VR equipment on anxiety about the test, the B value was 0.750 (P<0.01) and the t value was 6.181 (P<0.01). decreased and showed an influence of 75%. In addition, overall satisfaction with VR equipment was 3.76±1.28, and the intention to reuse was 66%. The Cronbach α value of the reliability coefficient of the questionnaire was 0.901. Conclusion When using VR equipment, patients' attention was dispersed, anxiety was reduced, and psychological stability was found. In the future, as VR equipment technology develops, it is thought that if the equipment can be miniaturized and the resolution of VR content images is increased, it can be used in various clinical settings if it provides more realistic stability to the patient.

• Journal of Korean Society of Forest Science
• /
• v.112 no.4
• /
• pp.426-441
• /
• 2023

This study investigated the predictive accuracy of a model of landslide displacement in Jecheon-si, where a great number of landslides were triggered by heavy rain on both natural (non-clear-cut) and clear-cut slopes during August 2020. This was accomplished by applying three flow direction methods (single flow direction, SFD; multiple flow direction, MFD; infinite flow direction, IFD) and the degree of root cohesion to an infinite slope stability equation. The application assumed that the soil saturation and any changes in root cohesion occurred following the timber harvest (clear-cutting). In the study area, 830 landslide locations were identified via landslide inventory mapping from satellite images and 25 cm resolution aerial photographs. The results of the landslide modeling comparison showed the accuracy of the models that considered changes in the root cohesion following clear-cutting to be improved by 1.3% to 2.6% when compared with those not considered in the area under the receiver operating characteristics (AUROC) analysis. Furthermore, the accuracy of the models that used the MFD algorithm improved by up to 1.3% when compared with the models that used the other algorithms in the AUROC analysis. These results suggest that the discriminatory application of the root cohesion, which considers changes in the vegetation condition, and the selection of the flow direction method may influence the accuracy of landslide predictive modeling. In the future, the results of this study should be verified by examining the root cohesion and its dynamic changes according to the tree species using the field hydrological monitoring technique.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_3
• /
• pp.979-995
• /
• 2023

As wildfires are difficult to predict, real-time monitoring is crucial for a timely response. Geostationary satellite images are very useful for active fire detection because they can monitor a vast area with high temporal resolution (e.g., 2 min). Existing satellite-based active fire detection algorithms detect thermal outliers using threshold values based on the statistical analysis of brightness temperature. However, the difficulty in establishing suitable thresholds for such threshold-based methods hinders their ability to detect fires with low intensity and achieve generalized performance. In light of these challenges, machine learning has emerged as a potential-solution. Until now, relatively simple techniques such as random forest, Vanilla convolutional neural network (CNN), and U-net have been applied for active fire detection. Therefore, this study proposed an active fire detection algorithm using state-of-the-art (SOTA) deep learning techniques using data from the Advanced Himawari Imager and evaluated it over East Asia and Australia. The SOTA model was developed by applying EfficientNet and lion optimizer, and the results were compared with the model using the Vanilla CNN structure. EfficientNet outperformed CNN with F1-scores of 0.88 and 0.83 in East Asia and Australia, respectively. The performance was better after using weighted loss, equal sampling, and image augmentation techniques to fix data imbalance issues compared to before the techniques were used, resulting in F1-scores of 0.92 in East Asia and 0.84 in Australia. It is anticipated that timely responses facilitated by the SOTA deep learning-based approach for active fire detection will effectively mitigate the damage caused by wildfires.

• Journal of Public Diplomacy
• /
• v.2 no.2
• /
• pp.27-52
• /
• 2022

Objectives: The neighboring countries of South Korea and Japan in Northeast Asia have interacted in both positive and negative ways, at times as close partners and other times adversaries, throughout their long and thorny history of extensive dynamics. The controversial dispute over Dokdo is one of the most critical issues evoking harsh tensions and arguments asserting wholly opposite claims. Dokdo is a small island between two coastal states, but significant in terms of territorial, botanical, and marine resources, and thus ownership of the island has become a point of conflict accompanied by a troubled history. But why has Dokdo been a source of conflicts and how should the controversial Dokdo issue be addressed in a way that fosters positive influence and co-prosperity? Methods: This study provides comprehensive and critical insights from a wealth of previous research and strategic suggestions for the Korean government. It utilizes the three perspectives of historical documents and political context, international regulations and legal frames, and public diplomacy. Furthermore, it applies these resources to negotiation theories and strategies to propose reasonable solutions. Results: This study suggests that it is important for Korea and Japan to try to build mutual trust through more active communication and interaction in order to understand each other before attempting to create a formal resolution via negotiation. In addition to these efforts, Korea needs to be ready for the inevitable need to take decisive action in terms of negotiation, using analytic and efficient strategies. The study proposes three solutions: 1) Strong Action Strategy, 2) International Legal Strategy, and 3) Public Diplomacy Strategy. Conclusions: From the perspective of public diplomacy, the Dokdo issue needs to be converted from a symbol of conflicts between Korea and Japan into a symbol of peace and co-prosperity. In addition to promoting a positive relationship between the two states, it can also contribute to the security environment of the Northeast Asian region and global peace.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.27 no.2
• /
• pp.99-109
• /
• 2023

Purpose: CT scan makes up for the weak point of the nuclear medicine image having a low resolution and also were used for attenuation correction on image reconstruction. Recently, many studies try to make use of CT images additionally, one of them is to measure the bone mineral density(BMD) using Quantitative CT(QCT) software. BMD exams are performed to scan lumbar and femur with DXA(Dual-Energy X-Ray Absorptiometry) in order to diagnose bone disease such as osteopenia, osteoporosis. The purpose of this study is to identify the usefulness of QCT_BMD analyzed with low dose CT images on L-spine Bone SPECT/CT comparing with DXA_BMD. Materials and Methods: Fifty five women over 50 years old (mean 66.4 ± 9.1) who took the both examinations(L-spine Bone SPECT/CT with SIEMENS Intevo 16 and DXA scan with GE Lunar prodigy advance) within 90 days from April 2017 to July 2022, BMD, T-score and disease classification were analyzed. Three-dimensional BMD was analyzed with low dose CT images acquired on L-spine Bone SPECT/CT scan on Mindways QCT PRO^TM software and two-dimensional BMD was analyzed on DXA scan. Basically, Lumbar 1-4 were analyzed and the patients who has lesion or spine implants on L-spine were excluded for this study. Pearson's correlation analysis was performed in BMD and T-score, chi-square test was performed in disease classification between QCT and DXA. Results: On 55 patients, the minimum of QCT_BMD was 18.10, maximum was 166.50, average was 82.71 ± 31.5 mg/cm³. And the minimum of DXA-BMD was 0.540, maximum was 1.302, average was 0.902 ± 0.201 g/cm², respectively. The result shows a strong statistical correlation between QCT_BMD and DXA_BMD(p<0.001, r=0.76). The minimum of QCT_T-score was -5.7, maximum was -0.1, average was -3.2 ± 1.3 and the minimum of DXA_T-score was -5.0, maximum was 1.7, average was -2.0 ± 1.3, respectively. The result shows a statistical correlation between QCT T-score and DXA T-score (p<0.001, r=0.66). On the disease classification, normal was 5, osteopenia was 25, osteoporosis was 25 in QCT and normal was 10, osteopenia was 25, osteoporosis was 20 in DXA. There was under-estimation of bone decrease relatively on DXA than QCT, but there was no significant differences statistically by chi-square test between QCT and DXA. Conclusion: Through this study, we could identify that the QCT measurement with low dose CT images QCT from L-Spine Bone SPECT/CT was reliable because of a strong statistical correlation between QCT_BMD and DXA_BMD. Bone SPECT/CT scan can provide three-dimensional information also BMD measurement with CT images. In the future, rather than various exams such as CT, BMD, Bone scan are performed, it will be possible to provide multipurpose information via only SPECT/CT scan. In addition, it will be very helpful clinically in the sense that we can provide a diagnosis of potential osteoporosis, especially in middle-aged patients.

• Journal of the Korean Society of Radiology
• /
• v.83 no.2
• /
• pp.344-359
• /
• 2022

Purpose To develop a denoising convolutional neural network-based image processing technique and investigate its efficacy in diagnosing breast cancer using low-dose mammography imaging. Materials and Methods A total of 6 breast radiologists were included in this prospective study. All radiologists independently evaluated low-dose images for lesion detection and rated them for diagnostic quality using a qualitative scale. After application of the denoising network, the same radiologists evaluated lesion detectability and image quality. For clinical application, a consensus on lesion type and localization on preoperative mammographic examinations of breast cancer patients was reached after discussion. Thereafter, coded low-dose, reconstructed full-dose, and full-dose images were presented and assessed in a random order. Results Lesions on 40% reconstructed full-dose images were better perceived when compared with low-dose images of mastectomy specimens as a reference. In clinical application, as compared to 40% reconstructed images, higher values were given on full-dose images for resolution (p < 0.001); diagnostic quality for calcifications (p < 0.001); and for masses, asymmetry, or architectural distortion (p = 0.037). The 40% reconstructed images showed comparable values to 100% full-dose images for overall quality (p = 0.547), lesion visibility (p = 0.120), and contrast (p = 0.083), without significant differences. Conclusion Effective denoising and image reconstruction processing techniques can enable breast cancer diagnosis with substantial radiation dose reduction.

• Korean Journal of Radiology
• /
• v.25 no.1
• /
• pp.11-23
• /
• 2024

Objective: To investigate whether reader training improves the performance and agreement of radiologists in interpreting unenhanced breast magnetic resonance imaging (MRI) scans using diffusion-weighted imaging (DWI). Materials and Methods: A study of 96 breasts (35 cancers, 24 benign, and 37 negative) in 48 asymptomatic women was performed between June 2019 and October 2020. High-resolution DWI with b-values of 0, 800, and 1200 sec/mm² was performed using a 3.0-T system. Sixteen breast radiologists independently reviewed the DWI, apparent diffusion coefficient maps, and T1-weighted MRI scans and recorded the Breast Imaging Reporting and Data System (BI-RADS) category for each breast. After a 2-h training session and a 5-month washout period, they re-evaluated the BI-RADS categories. A BI-RADS category of 4 (lesions with at least two suspicious criteria) or 5 (more than two suspicious criteria) was considered positive. The per-breast diagnostic performance of each reader was compared between the first and second reviews. Inter-reader agreement was evaluated using a multi-rater κ analysis and intraclass correlation coefficient (ICC). Results: Before training, the mean sensitivity, specificity, and accuracy of the 16 readers were 70.7% (95% confidence interval [CI]: 59.4-79.9), 90.8% (95% CI: 85.6-94.2), and 83.5% (95% CI: 78.6-87.4), respectively. After training, significant improvements in specificity (95.2%; 95% CI: 90.8-97.5; P = 0.001) and accuracy (85.9%; 95% CI: 80.9-89.8; P = 0.01) were observed, but no difference in sensitivity (69.8%; 95% CI: 58.1-79.4; P = 0.58) was observed. Regarding inter-reader agreement, the κ values were 0.57 (95% CI: 0.52-0.63) before training and 0.68 (95% CI: 0.62-0.74) after training, with a difference of 0.11 (95% CI: 0.02-0.18; P = 0.01). The ICC was 0.73 (95% CI: 0.69-0.74) before training and 0.79 (95% CI: 0.76-0.80) after training (P = 0.002). Conclusion: Brief reader training improved the performance and agreement of interpretations by breast radiologists using unenhanced MRI with DWI.

• Journal of Korean Dental Hygiene Science
• /
• v.1 no.2
• /
• pp.9-19
• /
• 2018

The purpose of this study was to examine the state of conflicts among dental health care workers. A survey was conducted on 266 dental hygienists and nursing assistants who worked in dental institutions from September 12 to November 13, 2017, and SPSS(statistical package for the social science) version 20.0 was employed to analyze the collected data. The findings of the study were as follows: 1. The most common reason of the dental hygienists for turnover was working hours and heavy workload(24.6%), followed by pay (22.6%), conflicts with dentists(16.0%) and conflicts with colleagues (11.3%). The most dominant reason of the nursing assistants for turnover was pay(31.1%), followed by working hours(24.4%), heavy workload(17.8%), conflicts with dentists(15.6%) and conflicts with colleagues(8.9%). 2. The largest reason for unsuccessful communication with dentists was that heavy workload reduced the opportunity to communicate well(54.5%). The second biggest reason was that they couldn't communicate well though they had the opportunity(24.0%), and the third greatest reason was that they tended to lag behind dentists in terms of professional knowledge(16.9%). 3. The biggest reason for unsuccessful communication among the dental health care workers was that they didn't have a lot of chances to communicate well on account of heavy workload(41.0%). The second largest reason was the differences in professional knowledge(24.9%), and the third greatest reason was that they couldn't communicate well though they had the chance(23.7%). 4. The most dominant reason for conflicts with dentists was the difference in power(24.0%), followed by poor communication skills(22.1%) and a lack of mutual respect(18.1%). But the opinions of the nursing assistants were different from those of the dental hygienists, as they cited poor communication skills as the most common reason, which was followed by the difference in power and a shortage of understanding of each other's work. 5. The most common reason for conflicts among the dental health care workers was a shortage of communication and communication skills(22.9%), and the second most dominant reasons were a lack of mutual respect and poor understanding of each other's work(17.5%), followed by a lack of mutual respect(17.2%). 6. As to the ways of resolving conflicts with dentists, the most common case was making some mutual concessions to compromise (28.9%), followed by delivering opinions through the staff meeting (23.9%), resolving conflicts by candidly exchanging opinions(15.8%), avoiding each other in moderation(11.7%) and following the opinions or assertions of dentists(1.3%). 7. Concerning the conflict resolution methods among the dental health care workers, the most prevalent way was making some mutual concessions to compromise(36.4%), followed by resolving conflicts by candidly exchanging opinions(23.0%) and conveying opinions through the staff meeting(18.5%). 8. Regarding communication among the dental health care workers, the dental hygienists(3.53±.729) considered themselves to be better at communicating than the nursing assistants(3.29±.745) did(p<0.05), and the dental hygienists(3.45±.809) who thought there was respectful treatment among workers who were different in occupational categories found themselves to be better than the nursing assistants(3.21±.952) who had the same thought did(p<0.05). As a result of analyzing whether frequent job-related meetings occurred among the workers whose occupational categories were different, the dental hygienists(3.05±.975) perceived that there were more frequent meetings than the nursing assistants(2.67±.955) did (p<0.01).

• Pediatric Infection and Vaccine
• /
• v.31 no.1
• /
• pp.37-45
• /
• 2024

Purpose: This study aimed to examine the clinical features and determinants of macrolide-unresponsive Mycoplasma pneumoniae pneumonia (MUMP) and to assess the differences in the time to fever resolution between doxycycline (DXC), tosufloxacin (TFX) and corticosteroid (CST) as second-line treatment. Methods: We retrospectively analyzed the medical records of patients under the age of 18 who were admitted to Nowon Eulji University Hospital between July 2018 and February 2020, diagnosed with mycoplasma pneumonia. Macrolide resistance was confirmed by detecting point mutations in the 23S rRNA gene. MUMP was clinically defined by persistent fever (≥38.0℃) lasting for 72 hours or more after the initiation of macrolide treatment. In cases of MUMP, patients were treated with an addition of CST, or the initial macrolide was replaced either DXC or TFX. Results: Out of 157 cases of mycoplasma pneumonia, 83 cases (52.9%) did not respond to macrolides. Patients with MUMP exhibited significantly higher C-reactive protein (CRP) levels (3.2±3.0 vs. 2.4±2.2 mg/dL, P=0.047), more frequent lobar/segmental infiltrations or pleural effusions (56.6% vs. 27.0%, P<0.001; 6.0% vs. 0.0%, P=0.032), and a higher prevalence of 23S rRNA gene mutations (96.4% vs. 64.6%, P<0.001) when compared to those with macrolide-susceptible M. pneumoniae pneumonia. In terms of second-line treatment, 15 patients (18.1%) responded to CST, 30 (36.1%) to DXC, and 38 (45.8%) to TFX. The time to defervescence (TTD) after initiation second-line treatment was significantly shorter in the CST group compared to the DXC (10.3±12.7 vs. 19.4±17.2 hours, P=0.003) and TFX groups (10.3±12.7 vs. 25.0±20.1 hours, P=0.043), with no significant difference observed between the DXC and TFX groups (19.4±17.2 vs. 25.0±20.1 hours, P=0.262). Conclusions: High CRP levels, the presence of positive 23S rRNA gene mutation, lobar or segmental lung infiltration, and pleural effusion observed in chest X-ray findings were significant factors associated with macrolide unresponsiveness. In this study, CST demonstrated a shorter TTD compared to DXC or TFX. Further, larger-scale prospective studies are needed to determine the optimal second-line treatment for MUMP.