• Tunnel and Underground Space
• /
• v.33 no.6
• /
• pp.508-518
• /
• 2023

This paper presents a new approach for the automatic mapping of discontinuities in a tunnel face based on its 3D digital model reconstructed by LiDAR scan or photogrammetry techniques. The main idea revolves around the identification of discontinuity areas in the 3D digital model of a tunnel face by segmenting its 2D projected images using a deep-learning semantic segmentation model called U-Net. The proposed deep learning model integrates various features including the projected RGB image, depth map image, and local surface properties-based images i.e., normal vector and curvature images to effectively segment areas of discontinuity in the images. Subsequently, the segmentation results are projected back onto the 3D model using depth maps and projection matrices to obtain an accurate representation of the location and extent of discontinuities within the 3D space. The performance of the segmentation model is evaluated by comparing the segmented results with their corresponding ground truths, which demonstrates the high accuracy of segmentation results with the intersection-over-union metric of approximately 0.8. Despite still being limited in training data, this method exhibits promising potential to address the limitations of conventional approaches, which only rely on normal vectors and unsupervised machine learning algorithms for grouping points in the 3D model into distinct sets of discontinuities.

• Phonetics and Speech Sciences
• /
• v.15 no.4
• /
• pp.109-118
• /
• 2023

Cardiac arrest is a critical medical emergency where immediate response is essential for patient survival. This is especially true for Out-of-Hospital Cardiac Arrest (OHCA), for which the actions of emergency medical services in the early stages significantly impact outcomes. However, in Korea, a challenge arises due to a shortage of dispatcher who handle a large volume of emergency calls. In such situations, the implementation of a machine learning-based OHCA detection program can assist responders and improve patient survival rates. In this study, we address this challenge by developing a machine learning-based OHCA detection program. This program analyzes transcripts of conversations between responders and callers to identify instances of cardiac arrest. The proposed model includes an automatic transcription module for these conversations, a text-based cardiac arrest detection model, and the necessary server and client components for program deployment. Importantly, The experimental results demonstrate the model's effectiveness, achieving a performance score of 79.49% based on the F1 metric and reducing the time needed for cardiac arrest detection by 15 seconds compared to dispatcher. Despite working with a limited dataset, this research highlights the potential of a cardiac arrest detection program as a valuable tool for responders, ultimately enhancing cardiac arrest survival rates.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.12
• /
• pp.505-518
• /
• 2023

3D point cloud semantic segmentation is a computer vision task that involves dividing the point cloud into different objects and regions by predicting the class label of each point. Existing 3D semantic segmentation models have some limitations in performing sufficient fusion of multi-modal features while ensuring both characteristics of 2D visual features extracted from RGB images and 3D geometric features extracted from point cloud. Therefore, in this paper, we propose MMCA-Net, a novel 3D semantic segmentation model using 2D-3D multi-modal features. The proposed model effectively fuses two heterogeneous 2D visual features and 3D geometric features by using an intermediate fusion strategy and a multi-modal cross attention-based fusion operation. Also, the proposed model extracts context-rich 3D geometric features from input point cloud consisting of irregularly distributed points by adopting PTv2 as 3D geometric encoder. In this paper, we conducted both quantitative and qualitative experiments with the benchmark dataset, ScanNetv2 in order to analyze the performance of the proposed model. In terms of the metric mIoU, the proposed model showed a 9.2% performance improvement over the PTv2 model using only 3D geometric features, and a 12.12% performance improvement over the MVPNet model using 2D-3D multi-modal features. As a result, we proved the effectiveness and usefulness of the proposed model.

• Convergence Security Journal
• /
• v.24 no.1
• /
• pp.59-68
• /
• 2024

With advancements in artificial intelligence technology, intelligent CCTV systems are being deployed across various environments, such as river bridges and construction sites. However, a conflict arises regarding the opening and closing of rooftop access points due to concerns over potential accidents and crime incidents and their role as emergency evacuation spaces. While the relevant law typically mandates the constant opening of designated rooftop access points, closures are often tacitly permitted in practice for security reasons, with a lack of appropriate legal measures. In this context, this study proposes a detection system utilizing intelligent CCTV to respond to emergencies that may occur on rooftops. We develop a system based on the YOLOv5 object detection model to detect assault and suicide attempts by jumping, introducing a new metric to assess them. Experimental results demonstrate that the proposed system rapidly detects assault and suicide attempts with high accuracy. Additionally, through a legal analysis of rooftop access point management, deficiencies in the legal framework regarding rooftop access and CCTV installation are identified, and improvement measures are proposed. With technological and legal improvements, we believe that crime and accident incidents in rooftop environments will decrease.

• Korean Journal of Radiology
• /
• v.25 no.5
• /
• pp.438-448
• /
• 2024

Objective: To evaluate the therapeutic outcomes of no-touch radiofrequency ablation (NT-RFA) using twin cooled wet (TCW) electrodes in patients experiencing recurrent hepatocellular carcinoma (HCC) after undergoing locoregional treatments. Materials and Methods: We conducted a prospective, single-arm study of NT-RFA involving 102 patients, with a total of 112 recurrent HCCs (each ≤ 3 cm). NT-RFA with TCW electrodes was implemented under the guidance of ultrasonography (US)-MR/CT fusion imaging. If NT-RFA application proved technically challenging, conversion to conventional tumor puncture RFA was permitted. The primary metric for evaluation was the mid-term cumulative incidence of local tumor progression (LTP) observed post-RFA. Cumulative LTP rates were estimated using the Kaplan-Meier method. Multivariable Cox proportional hazard regression was used to explore factors associated with LTP. Considering conversion cases from NT-RFA to conventional RFA, intention-to-treat (ITT; including all patients) and per-protocol (PP; including patients not requiring conversion to conventional RFA alone) analyses were performed. Results: Conversion from NT-RFA to conventional RFA was necessary for 24 (21.4%) out of 112 tumors. Successful treatment was noted in 111 (99.1%) out of them. No major complications were reported among the patients. According to ITT analysis, the estimated cumulative incidences of LTP were 1.9%, 6.0%, and 6.0% at 1, 2, and 3 years post-RFA, respectively. In PP analysis, the cumulative incidence of LTP was 0.0%, 1.3%, and 1.3% at 1, 2, and 3 years, respectively. The number of previous locoregional HCC treatments (adjusted hazard ratio [aHR], 1.265 per 1 treatment increase; P = 0.004), total bilirubin (aHR, 7.477 per 1 mg/dL increase; P = 0.012), and safety margin ≤ 5 mm (aHR, 9.029; P = 0.016) were independently associated with LTP in ITT analysis. Conclusion: NT-RFA using TCW electrodes is a safe and effective treatment for recurrent HCC, with 6.0% (ITT analysis) and 1.3% (PP analysis) cumulative incidence of LTP at 2 and 3-year follow-ups.

• Korean Journal of Radiology
• /
• v.24 no.5
• /
• pp.434-443
• /
• 2023

Objective: To investigate the association between pelvic bone computed tomography (CT)-derived body composition and patient outcomes in older adult patients who underwent surgery for proximal femur fractures. Materials and Methods: We retrospectively identified consecutive patients aged ≥ 65 years who underwent pelvic bone CT and subsequent surgery for proximal femur fractures between July 2018 and September 2021. Eight CT metrics were calculated from the cross-sectional area and attenuation of the subcutaneous fat and muscle, including the thigh subcutaneous fat (TSF) index, TSF attenuation, thigh muscle (TM) index, TM attenuation, gluteus maximus (GM) index, GM attenuation, gluteus medius and minimus (Gmm) index, and Gmm attenuation. The patients were dichotomized using the median value of each metric. Multivariable Cox regression and logistic regression models were used to determine the association between CT metrics with overall survival (OS) and postsurgical intensive care unit (ICU) admission, respectively. Results: A total of 372 patients (median age, 80.5 years; interquartile range, 76.0-85.0 years; 285 females) were included. TSF attenuation above the median (adjusted hazard ratio [HR], 2.39; 95% confidence interval [CI], 1.41-4.05), GM index below the median (adjusted HR, 2.63; 95% CI, 1.33-5.26), and Gmm index below the median (adjusted HR, 2.33; 95% CI, 1.12-4.55) were independently associated with shorter OS. TSF index (adjusted odds ratio [OR], 6.67; 95% CI, 3.13-14.29), GM index (adjusted OR, 3.45; 95% CI, 1.49-7.69), GM attenuation (adjusted OR, 2.33; 95% CI, 1.02-5.56), Gmm index (adjusted OR, 2.70; 95% CI, 1.22-5.88), and Gmm attenuation (adjusted OR, 2.22; 95% CI, 1.01-5.00) below the median were independently associated with ICU admission. Conclusion: In older adult patients who underwent surgery for proximal femur fracture, low muscle indices of the GM and gluteus medius/minimus obtained from their cross-sectional areas on preoperative pelvic bone CT were significant prognostic markers for predicting high mortality and postsurgical ICU admission.

• Korean Journal of Ichthyology
• /
• v.35 no.4
• /
• pp.285-293
• /
• 2023

This study conducted to investigate the temporal and spatial variations in the fish community structure in the Seomjin River estuary (ES1~ES3) and Gwangyang Bay (ES4~ES8). Surveys were conducted at eight points between May and September from 2020 to 2021. A total of 5,111 fish were collected, representing 42 families, 68 genera, and 78 species. The dominant species collected during the survey period were Pennahia argentata, with 2,370 individuals and a relative abundance of 46.4%. Also, as a subdominant species, Nuchequula nuchalis, 1643 individuals were caught, accounting for 32.1% relative abundance. According to Bray-Curtis similarities, observations were divided into three groups based on the fish community composition, and they were further categorized into brackish water areas, coastal areas, inner areas, and outer areas. Group 1 mainly consisted of fish inhabiting freshwater and brackish water areas, while group 3 exclusively contained fish living in marine waters. ANOSIM (Analysis of Similarity) analysis revealed a significant difference (ANOSIM R=0.783, p=0.001) between Group 2 and Group 3, and the species contributing to these clusters were P. argentata and Muraenesox cinereus. The study demonstrated spatial changes in the fish community structure from the Seomjin River estuary to the outer coast. The findings highlight the importance of long-term monitoring of major species that exhibit spatial variations, serving as essential foundational data for the conservation and management of economically significant fish species.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.11 no.3
• /
• pp.97-107
• /
• 2006

The nektonic assemblages of upper tidal flat area located in the Jangbong Island, near Incheon, were studied using a fence net from March to November 2001. A total of 49 species were recorded, with a mean abundance and biomass of 489 individuals and 5,170.4 g, respectively. The most abundant species by number were Exopalaemon carinicauda(40.9%) and Johnius grypotus(13.2%). By catch weight the dominant species were Acanthogobius hasta(33.7%), Johnius grypotus(14.6%) and Scomberomorus niphonius(10.2%). The conventional multivariate statistics(Cluster analysis and non-metric Multi-Dimensional Scaling) applied to assess temporal variation in nektonic communities. As a result of cluster analysis and MDS ordination, the faunal group could be divided into spring and summer/autumn dominant species group. The spring species included the pisces Acanthogobius luridus, the crab Macrophthalmus japonicus and the gastropods Bullacta exarata and Lunatia gilva. The summer/autumn species were the pisces J. grypotus, Sardinella zunasi, Konosirus punctatus, Chelon haematocheila, S. niphonius and Takifugu niphobles, the shrimp Metapenaeus joyneri and the cephalopod Loligo beka.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.28 no.12 s.138
• /
• pp.1583-1595
• /
• 2004

Simon and Sullivan(l993) estimated that clothing and textile related brand equity had the highest magnitude comparing any other industry category. It reflects that fashion brands reinforce the symbolic, social values and emotional characteristics being different from generic brands. Recently, Kim and Lim(2002) developed a fashion brand equity scale to measure a brand's psychometric properties. However, they suggested that additional psychometric tests were needed to compare the relative magnitude of each brand's equity. The purpose of this study was to recognize the psychometric constructs of fashion brand equity and validate Kim and Lim's fashion brand equity scale using the measurement invariance test of cross-group comparison. First, we identified the constructs of fashion brand equity using confirmatory factor analysis through structural equation modeling. Second, we compared the relative magnitude of two brands' equity using the measurement invariance test of multi-group simultaneous factor analysis. Data were collected at six major universities in Seoul, Korea. There were 696 usable surveys for data analysis. The results showed that fashion brand equity was comprised of 16 items representing six dimensions: customer-brand resonance, customer feeling, customer judgment, brand imagery, brand performance and brand awareness. Also, we could support the measurement invariance of two brands' equities by configural and metric invariance tests. There were significant differences in five constructs' mean values. The greatest difference was in customer feeling; the smallest, in customer judgment.

• Korean Journal of Environmental Biology
• /
• v.26 no.4
• /
• pp.311-322
• /
• 2008

The objectives of the study were to analyze chemical water quality and physical habitat characteristics in the urban streams (Miho and Gap streams) along with evaluations of fish community structures and ecosystem health, throughout fish composition and guild analyses during 2006$\sim$2007. Concentrations of BOD and COD averaged 3.5 and 5.7 mg L$^{-1}$, in the urban streams, while TN and TP averaged 5.1 mg L$^{-1}$ and 274 ${\mu}g$ L$^{-1}$, indicating an eutrophic state. Especially, organic pollution and eutrophication were most intense in the downstream reach of both streams. Total number of fish was 34 species in the both streams, and the most abundant species was Zacco platypus (32$\sim$42% of the total). In both streams, the relative abundance of sensitive species was low (23%) and tolerant and omnivores were high (45%, 52%), indicating an typical tolerance and trophic guilds of urban streams in Korea. According to multi-metric models of Stream Ecosystem Health Assessments (SEHA), model values were 19 and 24 in Miho Stream and Gap Stream, respectively. Habitat analysis showed that QHEI (Qulatitative Habitat Evaluation Index) values were 123 and 135 in the two streams, respectively. The minimum values in the SEHA and QHEI were observed in the both downstreams, and this was mainly attributed to chemical pollutions, as shown in the water quality parameters. The model values of SEHA were strongly correlated with conductivity (r=-0.530, p=0.016), BOD (r=-0.578, p< 0.01), COD (r=-0.603, p< 0.01), and nutrients (TN, TP: r>0.40, p<0.05). This model applied in this study seems to be a useful tool, which could reflect the chemical water quality in the urban streams. Overall, this study suggests that consistent ecological monitoring is required in the urban streams for the conservations along with ecological restorations in the degradated downstrems.