• 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.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.26 no.6
• /
• pp.157-166
• /
• 2022

Structural supervision on the construction site has been performed based on visual inspection, which is highly labor-intensive and subjective. In this study, the remote technique was developed to improve the efficiency of the measurements on rebar spacing using a 360° camera and reconstructed 3D models. The proposed method was verified by measuring the spacings in reinforced concrete structure, where the twelve locations in the construction site (265 m²) were scanned within 20 seconds per location and a total of 15 minutes was taken. SLAM, consisting of SIFT, RANSAC, and General framework graph optimization algorithms, produces RGB-based 3D and 3D point cloud models, respectively. The minimum resolution of the 3D point cloud was 0.1mm while that of the RGB-based 3D model was 10 mm. Based on the results from both 3D models, the measurement error was from 10.8% to 0.3% in the 3D point cloud and from 28.4% to 3.1% in the RGB-based 3D model. The results demonstrate that the proposed method has great potential for remote structural supervision with respect to its accuracy and objectivity.

• Korean Journal of Exercise Nutrition
• /
• v.25 no.1
• /
• pp.23-29
• /
• 2021

[Purpose] This preliminary study aimed to develop a regression model to estimate the non-exercise activity thermogenesis (NEAT) of Korean adults using various easy-to-measure dependent variables. [Methods] NEAT was measured in 71 healthy adults (male n = 29; female n = 42). Statistical analysis was performed to develop a NEAT estimation regression model using the stepwise regression method. [Results] We confirmed that age^A, weight^B, heart rate (HR)_average^C, weight × HR_average^D, weight × HR_sum^E, systolic blood pressure (SBP) × HR_rest^F, fat mass ÷ height^2G, gender × HR_average^H, and gender × weight × HR_sum^I were important variables in various NEAT activity regression models. There was no significant difference between the measured NEAT values obtained using a metabolic gas analyzer and the predicted NEAT. [Conclusion] This preliminary study developed a regression model to estimate the NEAT in healthy Korean adults. The regression model was as follows: sitting = 1.431 - 0.013 × (A) + 0.00014 × (D) - 0.00005 × (F) + 0.006 × (H); leg jiggling = 1.102 - 0.011 × (A) + 0.013 × (B) + 0.005 × (H); standing = 1.713 - 0.013 × (A) + 0.0000017 × (I); 4.5 km/h walking = 0.864 + 0.035 × (B) + 0.0000041 × (E); 6.0 km/h walking = 4.029 - 0.024 × (C) + 0.00071 × (D); climbing up 1 stair = 1.308 - 0.016 × (A) + 0.00035 × (D) - 0.000085 × (F) - 0.098 × (G); and climbing up 2 stairs = 1.442 - 0.023 × (A) - 0.000093 × (F) - 0.121 × (G) + 0.0000624 × (E).

• Journal of the Korean Society for Railway
• /
• v.16 no.1
• /
• pp.47-51
• /
• 2013

In this study, we carried out an experimental shield TBM excavation model test using a down-scale device in soft clay, to understand tunnel-face stability properties in relation to changes in slurry pressure. We performed five tests according to tunnel depth (0.5D, 0.75D, 1.0D, 1.25D, 1.5D), and compared theoretical tunnel-face pressure with model test results. The range in theoretical tunnel-face slurry pressure ($P_{min}{\leq}P_{slurry\;pressure}{\leq}P_{max}$), which is determined by earth pressure and water level, was very similar to the model test result. This result was due to the more isotropic condition of the soft clay ground, than of rocky ground.

• Nutrition Research and Practice
• /
• v.5 no.5
• /
• pp.464-470
• /
• 2011

Growing evidence suggests an elevated risk for colorectal neoplasia among individuals with low levels of vitamin D, the biological actions of which are mediated by the vitamin D receptor (VDR). To investigate the association among vitamin D status, VDR polymorphisms (FokI, and BsmI), and colorectal adenoma, we conducted a meta-analysis of nine studies of circulating levels of 25-hydroxyvitamin D (25(OH)D) and five studies of FokI or BsmI polymorphisms in relation to colorectal adenomas. Study-specific relative risks (RRs) and 95% confidence intervals (CIs) were pooled using a random-effects model. A total of 3398 coloreetal adenomas for 25(OH)D and 1754 colorectal adenomas for VDR were included in the meta-analysis. We identified a significant inverse association between colorectal adenoma (combined RR, 0.93; 95% CI, 0.87-0.98 per 10 ng/mL increase in 25(OH)D levels). When we examined FokI and BsmI polymorphisms in the meta-analysis, we found no association for either FokI (combined RR, 1.00; 95% CI, 0.95-1.06) or BsmI (combined RR, 0.99; 95% CI, 0.93-1.05) in the additive model. These data suggest an inverse association between circulating 25(OH)D levels and colorectal adenoma risk.

• Journal of Integrative Natural Science
• /
• v.4 no.3
• /
• pp.210-215
• /
• 2011

Holographic quantitative structure-activity relationships (HQSAR) is a useful tool to correlates structures with their biological activities. HQSAR is a two dimensional (2D) QSAR methodology, which generates QSAR equations through 2D fingerprint and correlates it with biological activity. Here, we report a 2D-QSAR model for a series of fifty-one 3,4-dihydroxychalcones derivatives utilizing HQSAR methodology. We developed HQSAR model with 6 optimum numbers of components (ONC), which resulted in cross-validated correlation coefficient ($q^2$) of 0.855 with 0.283 standard error of estimate (SEE). The non-cross-validated correlation coefficient (r2) with 0.966 indicates the model is predictive enough for analysis. Developed HQSAR model was binned in to a hologram length of 257. Atomic contribution map revealed the importance of dihydroxy substitution on phenyl ring.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.16 no.6
• /
• pp.507-520
• /
• 2014

In this study, the surface displacement was investigated according to the various depth of cover when the tunnel excavation equipment was used in a clay soil. For this the laboratory scaled model test was carried out using the soil sample similar to the in-situ conditions. We carried out four tests according to tunnel depth(1.5D, 2.0D, 2.5D, 3.0D). The distribution of impact due to tunnelling was quantitatively analyzed in the three-dimension by measuring the surface displacement. In addition, the pattern of surface displacements was figured out.

• Structural Engineering and Mechanics
• /
• v.67 no.5
• /
• pp.505-516
• /
• 2018

In this paper, the effects of particle size and model scale of concrete has been investigated on the failure mechanism of PFC2D numerical models under uniaxial compressive test. For this purpose, rectangular models with same particle sizes and different model dimensions, i.e., $3mm{\times}6mm$, $6mm{\times}12mm$, $12mm{\times}24mm$, $25mm{\times}50mm$ and $54mm{\times}108mm$, were prepared. Also rectangular models with dimension of $54mm{\times}108mm$ and different particle sizes, i.e., 0.27 mm, 0.47 mm, 0.67 mm, 0.87 mm, 1.07 mm, 1.87 mm and 2.27 mm were simulated using PFC2D and tested under uniaxial compressive test. Concurrent with uniaxial test, direct shear test was performed on the numerical models. Dimension of the models were $75{\times}100mm$. Two narrow bands of particles with dimension of $37.5mm{\times}20mm$ were removed from upper and lower of the model to supply the shear test condition. The particle sizes in the models were 0.47 mm, 0.57 mm, 0.67 mm and 0.77 mm. The result shows that failure pattern was affected by model scale and particle size. The uniaxial compressive strength and shear strength were increased by increasing the model scale and particle size.

• Journal of the Korean Institute of Navigation
• /
• v.1 no.1
• /
• pp.1-16
• /
• 1977

The aspects of Omega phase prediction are briefly reviewed, and Swanson's Model and Pierce's Model are presented. The equations for the Omega phase prediction and the most probable coefficients of the propagating equations are derived on the base of Pierce's Model by the least square method. The coefficients are calculated from the data which are the phase differences between the pairs of the Station A (Aldra, Norway), C(Haiku, Hawaii), and D(La Mour, North Dakota) observed at Busan Harbor of the South Coast of Korea in June and September, 1976. It is clearly shown that the standard deviations of the observed lane values at Busan Harbor are as followed: 1. June, 1976. Pair (A-C) : 0.1446 Pair (C-D) : 0.2598 2.September, 1976. Pair (A-D) : 0.3958 Pafr (C-D) : 0.3278 As a conclusion of the above investigation, it is shown that the Omega phase velocity can be predicted by the method, proposed in this paper, of analyzing the diurnal and seasonal variations of the Omege phase velocity except SID, PCD and AZD. If more observed data are employed, more exact Omega phase velocity is expected to be obtained.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.6
• /
• pp.259-268
• /
• 2007

This paper contributes to an improvement of the statistical bandwidth extension(BWE) system based on Hidden Markov Model(HMM). First, the existing HMM training method for BWE, which is suggested originally by Jax, is analyzed in comparison with the general Baum-Welch training method. Next, based on this analysis, a new HMM-based BWE method is suggested which adopts the Baum-Welch re-estimation algorithm instead of the Jax's to train HMM model. Conclusionally speaking, the Baum-Welch re-estimation algorithm is a generalized form of the Jax's training method. It is flexible and adaptive in modeling the statistical characteristic of training data. Therefore, it generates a better model to the training data, which results in an enhanced BWE system. According to experimental results, the new method performs much better than the Jax's BWE systemin all cases. Under the given test conditions, the RMS log spectral distortion(LSD) scores were improved ranged from 0.31dB to 0.8dB, and 0.52dB in average.