• Journal of the Korea Society of Computer and Information
• /
• v.28 no.11
• /
• pp.1-11
• /
• 2023

Breast cancer is the disease that affects women the most worldwide. Due to the development of computer technology, the efficiency of machine learning has increased, and thus plays an important role in cancer detection and diagnosis. Deep learning is a field of machine learning technology based on an artificial neural network, and its performance has been rapidly improved in recent years, and its application range is expanding. In this paper, we propose a DNN-SVM hybrid model that combines the structure of a deep neural network (DNN) based on transfer learning and a support vector machine (SVM) for breast cancer classification. The transfer learning-based proposed model is effective for small training data, has a fast learning speed, and can improve model performance by combining all the advantages of a single model, that is, DNN and SVM. To evaluate the performance of the proposed DNN-SVM Hybrid model, the performance test results with WOBC and WDBC breast cancer data provided by the UCI machine learning repository showed that the proposed model is superior to single models such as logistic regression, DNN, and SVM, and ensemble models such as random forest in various performance measures.

• Korean Journal of Radiology
• /
• v.25 no.2
• /
• pp.189-198
• /
• 2024

Objective: To investigate the prognostic utility of radiomics features extracted from ¹⁸F-fluorodeoxyglucose (FDG) PET/CT combined with clinical factors and metabolic parameters in predicting progression-free survival (PFS) and overall survival (OS) in individuals diagnosed with extranodal nasal-type NK/T cell lymphoma (ENKTCL). Materials and Methods: A total of 126 adults with ENKTCL who underwent ¹⁸F-FDG PET/CT examination before treatment were retrospectively included and randomly divided into training (n = 88) and validation cohorts (n = 38) at a ratio of 7:3. Least absolute shrinkage and selection operation Cox regression analysis was used to select the best radiomics features and calculate each patient's radiomics scores (RadPFS and RadOS). Kaplan-Meier curve and Log-rank test were used to compare survival between patient groups risk-stratified by the radiomics scores. Various models to predict PFS and OS were constructed, including clinical, metabolic, clinical + metabolic, and clinical + metabolic + radiomics models. The discriminative ability of each model was evaluated using Harrell's C index. The performance of each model in predicting PFS and OS for 1-, 3-, and 5-years was evaluated using the time-dependent receiver operating characteristic (ROC) curve. Results: Kaplan-Meier curve analysis demonstrated that the radiomics scores effectively identified high- and low-risk patients (all P < 0.05). Multivariable Cox analysis showed that the Ann Arbor stage, maximum standardized uptake value (SUVmax), and RadPFS were independent risk factors associated with PFS. Further, β2-microglobulin, Eastern Cooperative Oncology Group performance status score, SUVmax, and RadOS were independent risk factors for OS. The clinical + metabolic + radiomics model exhibited the greatest discriminative ability for both PFS (Harrell's C-index: 0.805 in the validation cohort) and OS (Harrell's C-index: 0.833 in the validation cohort). The time-dependent ROC analysis indicated that the clinical + metabolic + radiomics model had the best predictive performance. Conclusion: The PET/CT-based clinical + metabolic + radiomics model can enhance prognostication among patients with ENKTCL and may be a non-invasive and efficient risk stratification tool for clinical practice.

• Korean Journal of Radiology
• /
• v.24 no.4
• /
• pp.274-283
• /
• 2023

Objective: To compare the outcomes of digital breast tomosynthesis (DBT) screening combined with ultrasound (US) with those of digital mammography (DM) combined with US in women with dense breasts. Materials and Methods: A retrospective database search identified consecutive asymptomatic women with dense breasts who underwent breast cancer screening with DBT or DM and whole-breast US simultaneously between June 2016 and July 2019. Women who underwent DBT + US (DBT cohort) and DM + US (DM cohort) were matched using 1:2 ratio according to mammographic density, age, menopausal status, hormone replacement therapy, and a family history of breast cancer. The cancer detection rate (CDR) per 1000 screening examinations, abnormal interpretation rate (AIR), sensitivity, and specificity were compared. Results: A total of 863 women in the DBT cohort were matched with 1726 women in the DM cohort (median age, 53 years; interquartile range, 40-78 years) and 26 breast cancers (9 in the DBT cohort and 17 in the DM cohort) were identified. The DBT and DM cohorts showed comparable CDR (10.4 [9 of 863; 95% confidence interval {CI}: 4.8-19.7] vs. 9.8 [17 of 1726; 95% CI: 5.7-15.7] per 1000 examinations, respectively; P = 0.889). DBT cohort showed a higher AIR than the DM cohort (31.6% [273 of 863; 95% CI: 28.5%-34.9%] vs. 22.4% [387 of 1726; 95% CI: 20.5%-24.5%]; P < 0.001). The sensitivity for both cohorts was 100%. In women with negative findings on DBT or DM, supplemental US yielded similar CDRs in both DBT and DM cohorts (4.0 vs. 3.3 per 1000 examinations, respectively; P = 0.803) and higher AIR in the DBT cohort (24.8% [188 of 758; 95% CI: 21.8%-28.0%] vs. 16.9% [257 of 1516; 95% CI: 15.1%-18.9%; P < 0.001). Conclusion: DBT screening combined with US showed comparable CDR but lower specificity than DM screening combined with US in women with dense breasts.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.37 no.4
• /
• pp.394-399
• /
• 2024

An irradiator is developed using two UVA wavelength ranges of SMD LEDs as a curing light source. This module has dimensions of 545×111×300 mm³ and is equipped with a TIR bar-shaped lens made of PDMS silicone resin. The developed irradiator offers high uniformity, with 89% in the centerline of the horizontal axis direction, for two different wavelength ranges of 365 nm and 385 nm. The radiation intensity from the light source module shows highly directional characteristics, and the irradiator provides a maximum irradiance of 1,634 mW/cm² at a working distance of 50 mm. During the initial 5 minutes of operation, the irradiance experiences a rapid decrease. However, this issue is addressed by optimizing the LED's current reduction characteristics and managing the Transistor's temperature rise in the constant current circuit. After continuous operation for approximately 60 minutes. The highest temperature, near the central part of the irradiating surface, reaches 69.7℃, while the lowest temperature, near the edges, is 41.1℃.

• Journal of Korea Water Resources Association
• /
• v.57 no.2
• /
• pp.73-85
• /
• 2024

Predicting water quality of rivers and reservoirs is necessary for the management of water resources. Artificial Neural Networks (ANNs) have been used in many studies to predict water quality with high accuracy. Previous studies have used Gradient Descent (GD)-based optimizers as an optimizer, an operator of ANN that searches parameters. However, GD-based optimizers have the disadvantages of the possibility of local optimal convergence and absence of a solution storage and comparison structure. This study developed improved optimizers to overcome the disadvantages of GD-based optimizers. Proposed optimizers are optimizers that combine adaptive moments (Adam) and Nesterov-accelerated adaptive moments (Nadam), which have low learning errors among GD-based optimizers, with Harmony Search (HS) or Novel Self-adaptive Harmony Search (NSHS). To evaluate the performance of Long Short-Term Memory (LSTM) using improved optimizers, the water quality data from the Dasan water quality monitoring station were used for training and prediction. Comparing the learning results, Mean Squared Error (MSE) of LSTM using Nadam combined with NSHS (NadamNSHS) was the lowest at 0.002921. In addition, the prediction rankings according to MSE and R² for the four water quality indices for each optimizer were compared. Comparing the average of ranking for each optimizer, it was confirmed that LSTM using NadamNSHS was the highest at 2.25.

• Journal of the Korea Concrete Institute
• /
• v.21 no.3
• /
• pp.245-253
• /
• 2009

The importance of chloride ions in the corrosion of steel in concrete has led to the concept for chloride threshold level (CTL). The CTL can be defined as the content of chlorides at the steel depth that is necessary to sustain local passive film breakdown and hence initiate the corrosion process. Despite the importance of the CTL, due to the uncertainty determining the actual limits in various environments for chloride-induced corrosion, conservative values such as 0.4% by weight of cement or 1.2 kg in 1 $m^3$ concrete have been used in predicting the corrosion-free service life of reinforced concrete structures. The paper studies the CTL for blended cement concrete by comparing the resistance of cementitious binder to the onset of chloride-induced corrosion of steel. Mortar specimens were cast with centrally located steel rebar of 10 mm in diameter using cementitious mortars with ordinary Portland cement (OPC) and mixed mortars replaced with 30% pulverized fuel ash (PFA), 60% ground granulated blast furnace slag (GGBS) and 10% silica fume (SF), respectively, at 0.4 of a free W/B ratio. Chlorides were admixed in mixing water ranging 0.0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.5, 2.0, 2.5 and 3.0% by weight of binder(Based on $C1^-$). Specimens were curd 28 days at the room temperature, wrapped in polyethylene film to avoid leaching out of chloride and hydroxyl ions. Then the corrosion rate was measured using the polarization resistance method and the order of CTL for binder was determined. Thus, CTL of OPC, 60%GGBS, 30%PFA and 10%SF were determined by 1.6%, 0.45%, 0.8% and 2.15%, respectively.

• Journal of dental hygiene science
• /
• v.11 no.3
• /
• pp.229-234
• /
• 2011

This study obtained DMFT index(9271 people) in population with more than 6 years old based on data of 2009 Korea National Health and Nutrition Examination Survey, and calculated SiC index(2,517 people) by classifying group with high risk, which is 1/3 in high rank out of it. The aim was to confirm existence of the group with high risk of dental caries, and was to emphasize necessity of preventive strategy for the group with high risk. Statistical analysis was carried out t-test by using STATA 11.0 program. 1. DMFT index accounted for 45.0%(4,174 people) for men and 55.0%(5,097 people) for women. SiC index stood at 2,517 people. Men accounted for 36.3%(914 people). Thus, women were larger with 64.7%(1,603 people). 2. Seeing the mean in DMFT index and SiC index, the more rise in age led to the bigger difference in the mean between two groups. Especially, the difference was shown more than 4 points on the average for 13 and 14 years old and more than double on the average from over 15 years. 3. As a result of comparing gender, DMFT index was higher in women than men from over 14 years old. SiC index wasn't significant. However, women were indicated to be 1 piece more than men at the age of 9 and 14. 4. As a result of comparing by region, both DMFT index and SiC index were indicated to be higher in the rural area for their over 60s. Seeing the above results, the group with high risk is showing the higher rate of experiencing dental caries than DMFT-index in the whole. Thus, the comprehensively & intensively preventive management seems to be likely needed on group with high risk.

• The Journal of Society for e-Business Studies
• /
• v.17 no.1
• /
• pp.23-38
• /
• 2012

Massive Multiplayer Online Role Playing Games(MMORPGs) headed by some Korean game companies such as NC Soft, NHN, and Nexon have exploded in recent years. However, it becomes one of the major challenges for the MMORPG developers to design their games to appeal to gamers since only a few MMORPGs succeed whereas they require a huge amount of initial investment. Under this background, our study derives the major elements for designing MMORPG from the literature, and identifies the ones critical to the users' satisfaction and their willingness to pay among the derived elements. Though most previous studies on the design elements of MMORPG have used analytic hierarchy process(AHP), our study adopts artificial neural network(ANN) as the tool for identifying key success factors in designing MMORPG. The results of our study show that the elements of the game contents quality have a bigger effect on the user's satisfaction, whereas the ones of the value-added systems have a bigger effect on the user's willingness to pay. They also show that user interface affects both the user's satisfaction and willingness to pay most. These results imply that the strategies for the development of MMORPG should be aligned with its goal and market penetration strategy. They also imply that the satisfaction and revenue generation from MMORPG cannot be achieved without convenient and easy control environment. It is expected that the new findings of our study would be useful forthe developers or publishers of MMORPGs to build their own business strategies.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.34 no.1
• /
• pp.13-20
• /
• 2005

The effects of pectin on obesity was studied using 3T3-L1 pre-adipocytes and rats fed 20% high fat diets. The concentration of leptin released from 3T3-L1 adipocytes in the presence of pectin was significantly decreased by 85% compared to that of the control (p<0.05), however, glycerol concentration was not changed. These data indicate that pectin seems to inhibit lipids accumulation in the adipocytes rather than enhance the lipolytic activity. Forty Sprague Dawley rats were fed 20% high fat diet for 8 weeks to induce obesity and then divided equally into four groups. Experimental groups were normal diet group (ND), high fat diet group (HFD), HDF with 10% pectin group (HFP10), and HDF with 20% pectin group (HFP20). Diet for the each group was prepared to be iso-caloric following AIN-76 guideline. After obesity was induced, rats were placed on an restricted diet for 9 weeks. The body weight of HFD increased 50% (p<0.05) compared to the ND, while it was decreased by 12% and 16% for HFP10 and HFP20, respectively (p<0.05). The relative amount of visceral fats for HFDl0 and HFD20 were decreased by 45% and 59% compared to that of HDF (130%), respectively (p<0.05). Pectin seems to have a greater effect on reducing visceral fats accumulation than weight reduction. Significantly increased level of triglyceride, total cholesterol or LDL-cholesterol in the plasma of HFD was returned to the normal or even below the normal by pectin diet, while the level of HDL-cholesterol increased. Lipid lowering effect was also observed in the liver and heart. These effects of pectin were dosedependent. In conclusion, the beneficial effect of pectin on the obesity was observed from cell culture experiment and animal study in terms of inhibiting the accumulation of lipids in the adipocytes.

• Horticultural Science & Technology
• /
• v.33 no.2
• /
• pp.219-226
• /
• 2015

The objective of this research was to investigate the influence of various root media on the growth of mother and daughter plants during propagation and early yield after transplanting of 'Seolhyang' strawberry. To achieve this, daughter plants were fixed to connected small pots that contained expanded rice-hull (ERH), a strawberry-specialized commercial medium (SSCM), soil mother materials (SMM), or loamy sand (LS). Then, growth of daughter plants in above- and below-ground tissue as well as early yield after transplanting to plastic house soil were investigated. The growth of daughter plants in terms of plant height, leaf area and fresh weight were the highest in the SSCM treatment. Root growth in terms of the amount of primary roots and root dry weight were the highest in the treatments of ERH and SMM and the lowest in that of SSCM, among treatments tested. The ERH treatment also showed the highest values among treatments in root length, surface area and volume when roots with 0 to 0.4 mm in diameter were investigated. The flower bud differentiation of daughter plants began on Sept. 3 in the ERH treatments, earlier than the SMM (Sept. 5) and in SSCM (Sept. 7) treatments. The tissue N contents of daughter plants were in the range of 1.41 to 1.55% in all treatments, and no significant differences were observed among treatments. This indicates that the low moisture retention capacity of ERH and water stress, rather than tissue N contents, promote the flower differentiation of daughter plants. In the evaluation of early yield after transplant, the ERH treatment of showed the highest yield in the period from November to December, reaching 667 g fruit weight per 10 plants. The yields per 10 plants in the other treatments were 581 g in SMM, 475 g in SSCM and 295 g in LS. Above results imply that the various root media have different effects on the growth of daughter plants as well as flower bud differentiation. Therefore, improvement in early yield after transplant can be achieved through selection of proper root medium for daughter plant propagation.