• Journal of Biomedical Engineering Research
• /
• v.43 no.1
• /
• pp.19-26
• /
• 2022

The Comprehensive in vitro Proarrhythmic Assay(CiPA) project was launched for solving the hERG assay problem of being classified as high-risk groups even though they are low-risk drugs due to their high sensitivity. CiPA presented a protocol to predict drug toxicity using physiological data calculated based on the in-silico model. in this study, features calculated through the in-silico model are analyzed for correlation of changing action potential in the near future, and features are verified through predictive performance according to drug datasets. Using the O'Hara Rudy model modified by Dutta et al., Pearson correlation analysis was performed between 13 features(dVm/dtmax, APpeak, APresting, APD90, APD50, APDtri, Ca_peak, Ca_resting, CaD90, CaD50, CaDtri, qNet, qInward) calculated at 100 pacing, and between dVm/d_{tmax_repol} calculated at 1,000 pacing, and linear regression analysis was performed on each of the 12 training drugs, 16 verification drugs, and 28 drugs. Indicators showing high coefficient of determination(R²) in the training drug dataset were qNet 0.93, AP resting 0.83, APDtri 0.78, Ca resting 0.76, dVm/dt_max 0.63, and APD90 0.61. The indicators showing high determinants in the validated drug dataset were APDtri 0.94, APD90 0.92, APD50 0.85, CaD50 0.84, qNet 0.76, and CaD90 0.64. Indicators with high coefficients of determination for all 28 drugs are qNet 0.78, APD90 0.74, and qInward 0.59. The indicators vary in predictive performance depending on the drug dataset, and qNet showed the same high performance of 0.7 or more on the training drug dataset, the verified drug dataset, and the entire drug dataset.

• Journal of Korean Neurosurgical Society
• /
• v.66 no.5
• /
• pp.582-590
• /
• 2023

Objective : Trauma is a leading causes of death and disability in all ages. The aim of this study was to describe the demography and characteristics of paediatric head trauma in our institution and examine the predictors of outcome and incidence of injury related mortality. Methods : We examined our institutional Trauma Registry over a 2 year period. Results : A total of 1100 trauma patients were seen over the study period. Of the 579 patients who had head injury 99 were in the paediatric age group. Of the paediatric head trauma patients 79 had documented Glasgow coma score (GCS), 38 (48.1%), 17 (21.5%) and 24 (30.4%) had mild, moderate and severe head injury respectively. The percentage mortality of head injury in the paediatric age group was 6.06% (6/99). There is an association between mortality and GCS (p=0.008), necessity for intensive care unit (ICU) admission (p=0.0001), associated burns (p=0.0001) and complications such as aspiration pneumonia (p=0.0001). The significant predictors of outcome are aspiration (p=0.004), the need for ICU admission (p=0.0001) and associated burns (p=0.005) using logistic binary regression. During the study period 46 children underwent surgical intervention with extradural haematoma 16 (34.8%), depressed skull fracture 14 (30.4%) and chronic subdural haematoma five (10.9%) being the commonest indication for surgeries. Conclusion : Paediatric head injury accounted for 9.0% (99/1100) of all trauma admissions. Majority of patients had mild or moderate injuries. Burns, aspiration pneumonitis and the need for ICU admission were important predictors of outcome in children with traumatic brain injury.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.10
• /
• pp.537-544
• /
• 2024

The rapid advancement of artificial intelligence and machine learning technologies is driving innovation across various industries, with natural language processing offering substantial opportunities for the analysis and processing of text data. The development of effective text classification models requires several complex stages, including data exploration, preprocessing, feature extraction, model selection, hyperparameter optimization, and performance evaluation, all of which demand significant time and domain expertise. Automated machine learning (AutoML) aims to automate these processes, thus allowing practitioners without specialized knowledge to develop high-performance models efficiently. However, current AutoML frameworks are primarily designed for structured data, which presents challenges for unstructured text data, as manual intervention is often required for preprocessing and feature extraction. To address these limitations, this study proposes a web-based AutoML platform that automates text preprocessing, word embedding, model training, and evaluation. The proposed platform substantially enhances the efficiency of text classification workflows by enabling users to upload text data, automatically generate the optimal ML model, and visually present performance metrics. Experimental results across multiple text classification datasets indicate that the proposed platform achieves high levels of accuracy and precision, with particularly notable performance when utilizing a Stacked Ensemble approach. This study highlights the potential for non-experts to effectively analyze and leverage text data through automated text classification and outlines future directions to further enhance performance by integrating Large language models.

• Korean Journal of Clinical Laboratory Science
• /
• v.49 no.2
• /
• pp.141-149
• /
• 2017

To examine the cause of acute diarrheal disease in children and adolescents, 521 fecal samples underwent multiple-PCR for a period of two years, between March 2015 and February 20178, in the Jeju region of Korea. Based on the analysis of 179 positive samples, 102 samples were male (56.98%) and 77 were female (43.02%), and highest positive rates were shown in the age group of 3~4 years (51.96%) and 5~6 years (12.29%). When 209 cases (40.12%), including double infection were analyzed, there were 88 cases (16.89%) of norovirus-GII infection, 26 cases (4.99%) of Campylobacter spp. infection, 18 cases (3.45%) of rotavirus infection, 18 cases (3.45%) of Clostridium difficile Toxin B infection, 17 cases (3.26%) of adenovirus infection, 16 cases (3.07%) of Clostridium perfringens infection, 11 cases (2.11%) of astrovirus infection, 5 cases (0.96%) of Salmonella spp. infection, 3 cases (0.58%) of norovirus-GI, Yersinia spp. and Aeromonas spp. infections, and 1 case (0.19%) of verocytotoxin-producing E. coli infection. Based on a seasonal separation of early childhood, norovirus-GI and norovirus-GII mainly prevailed during the winter, when the temperature is low. Typical enteritis with an increased prevalence of rotavirus during the spring. Astrovirus prevailed between the months of April and October, when norovirus-GI, norovirus-GII, and rotavirus did not prevail. With increasing age, acute diarrheal disease was not only induced by a virus, but also by bacteria. Although a test for virus is an effective method when trying to identify the cause during early childhood by multiplex-PCR, it would be desirable to undergo tests for both virus and bacteria concurrently as age increases.

• Journal of Life Science
• /
• v.28 no.7
• /
• pp.841-848
• /
• 2018

The flavonoid content and antioxidant effects of extracts from Stachys sieboldii Miq. and Lycopus lucidus Turcz were compared. The flavonoid content of the acetone + methylene chloride (A+M) extract of L. lucidus Turcz was 233.2 mg/g, suggesting that the extract was greater than that of S. sieboldii Miq. In the DPPH assay and the A+M and methanol (MeOH) extracts from L. lucidus Turcz had greater scavenging effects than those of S. sieboldii Miq. (p<0.05). The A+M extract from L. lucidus Turcz (0.5 mg/ml concentration) had an 82% scavenging effect in the DPPH assay. In the ABTS assay, A+M extracts from both S. sieboldii Miq. and L. lucidus Turcz (0.5 mg/ml concentration) had scavenging effects of 90% and 88%, respectively (p<0.05), suggesting that both A+M extracts had greater scavenging effects than those of both MeOH extracts. In a 120 min ROS production assay, all tested extracts dose-dependently decreased the cellular ROS production that was induced by $H_2O_2$, as compared to those produced by exposure to the extract-free control. The A+M extracts from both S. sieboldii Miq. and L. lucidus Turcz had greater inhibitory effects on cellular ROS production than those of both MeOH extracts at all concentrations tested. Treatment with the A+M extracts from S. sieboldii Miq. and L. lucidus Turcz (0.25 mg/ml concentration) inhibited the cellular ROS production by 60% and 86%, respectively. These results suggest that the A+M extracts of Stachys sieboldii Miq. and L. lucidus Turcz inhibit cellular oxidation and may contain valuable bioactive compounds, such as flavonoids.

• Korean Journal of Soil Science and Fertilizer
• /
• v.47 no.6
• /
• pp.579-585
• /
• 2014

This study was conducted to investigate liquid pig manure (LPM) application rates on the growth characteristics, yield, and feed value of whole crop barley in Yeongsangang and Saemangeum reclaimed tidal land. Electronic conductivity (EC), organic matter (OM), and available phosphate (Av. $P_2O_5$) increased in chemical properties of Yeongsangang and Saemangeum soil as raising LPM application level. As increasing LPM application level, exchangeable $Na^+$ significantly increased in Yeongsangang, while exchangeable $K^+$ significantly increased in Saemangeum. Plant height was not significantly different from LPM 100% to LPM 200% in Yeongsangang and in Saemangeum. Dry matter yield of whole crop barley increased steadily, but crop yield of LPM 200% in Yeongsangang ($10.5ton\;ha^{-1}$) was as much as that of LPM 150% ($10.0ton\;ha^{-1}$). Yield of LPM 200% ($11.2ton\;ha^{-1}$) in Saemangeum was similar to that of LPM 150% ($10.5ton\;ha^{-1}$). Crude protein (CP) increased depending on LPM application level, but total digestible nutrients (TDN) increased regardless of LPM application level. LPM 200% was the highest in TDN yield (Yeongsangang: $7.4ton\;ha^{-1}$, Saemangeum: $6.9ton\;ha^{-1}$), but there was no statistical difference between LPM 150% (Yeongsangang: $6.9ton\;ha^{-1}$, Saemangeum: $6.6ton\;ha^{-1}$) and LPM 200%. From the results described above, optimum rate of LPM for cultivating whole crop barley is considered 100% in Yeongsangang reclaimed tidal land and 150% in Saemangeum reclaimed tidal land, showing that the effect of LPM application is better in Segmentation than that in Yeongsangang for yield of whole crop barley.

• Journal of Dental Rehabilitation and Applied Science
• /
• v.28 no.2
• /
• pp.119-126
• /
• 2012

State of problem: Cement-retained implant-supported prostheses are routinely used in dentistry. The use of high strength cements has become more popular with the increasing confidence in the stability of the implant-abutment screw connection and the high survival rates of osseointegrated implants. No clinical data on retention of metal copings using CAD/CAM. To evaluate retention of metal copings using CAD/CAM system bonded to short titanium abutment with four different cements and compare retentive strength of metal copings with sandblasting or without sandblasting before cementation. Forty titanium abutment blocks were fabricated and divided into 4 groups of 10 samples each. Forty metal copings with occlusal hole to allow for retention testing were fabricated using CAD/CAM technology. The four cements were Fujicem(Fuji, Japan), Maxcem Elite(Kerr, USA), Panavia F2.0(Kurarary, Japan) and Superbond C&B(Sunmedical, Japan). The copings were cemented on the titanium abutment according to manufacture's recommendation. All samples were stored for 24h at 37oC in 100% humidity and tested for retention using universal testing machine(Instron) at a crosshead speed of 1.0mm/min. Force at retentive failure was recorded in Newton. The mode of failure was also recorded. Means and standard deviations of loads at failure were analyzed using ANOVA and Paired t-test. Statistical significance was set at P<0.05. Panavia F2.0 provided significantly higher retentive strength than Fujicem, Maxcem Elite(P<0.05). Sandblasting significantly increased bond strength(P<0.05). The mode of failure was cement remaining principally on metal copings. Within the limitation of this study, Panavia F2.0 showed significantly stronger retentive strength than Fujicem, Maxcem Elite(p<0.05). The Ranking order of the cements to retain the copings was Panavia F2.0, Fujicem = Maxcem Elite. Sandblasting significantly increased bond strength(P<0.05). The retentive strength of metal copings on implant abutment were influenced by surface roughness and type of cements.

• Tuberculosis and Respiratory Diseases
• /
• v.49 no.5
• /
• pp.614-623
• /
• 2000

Background : Diffuse alveolar hemorrhage (DAH) is rare but often fatal. To determine the clinical manifestations of DAH, its etiology, clinical course and prognosis were studied. Method : A retrospective analysis was performed in 21 patients that were diagnosed as DAH. Diagnosis of DAH was based on the presence of the "classical triad" of hemoptysis, anemia, and rapidly progressive infiltrates on chest X-ray and a finding of bronchoalveolar lavage or lung biopsy. Results : Thirteen patients (61.9%) had collagen vascular diseases (CVDs) as underlying disease and 10 patients had systemic lupus erythematosus. Females were more prevalent in CVD than in non-collagen vascular disease (NCVD). Otherwise, there were no significant differences between the two groups in terms of clinical manifestations. Dyspnea (95.2%), cough (76.2%), hemoptysis (61.9%), and fever (33.0%) were frequent symptoms. The initial creatinine level was higher in CVD than in NCVD ($3.27{\pm}3.15$ mg/dl vs. $1.19{\pm}0.94$ mg/dl, p=0.030). The corresponding drop in hemoglobin level was $2.69{\pm}1.26$ g/dl. Maximal drop in hemoglobin preceded the progression of infiltrates on the chest radiograph by $1.38{\pm}4.22$ days. The mortality rate was higher in the patients with NCVD than in those with CVD (50.0% vs. 23.1%). Conclusion : The DAH can occur not only in patients with CVD but also in those with NCVD. Higher creatinine level CVD in patients is associated with renal involvement in conjunction with DAH. The maximal drop in hemoglobin preceeding the progression of infiltrates on the chest radiograph suggests that the drop in hemoglobin is important for diagnosing DAH.

• Applied Biological Chemistry
• /
• v.8
• /
• pp.87-93
• /
• 1967

1) For the micro-analysis of mercury in plant materials, the method of Furutani was shown to be the simplest and most efficient way and the recovery of the assay was about 98%. 2) When the rice grain was soaked in 1/1000 diluted solution of organo-mercury fungicide for 8 hours at the end of March, the amounts of mercury residues in the brown rice and unhulled rice were 8.8 to $9.5\;{\mu}g/g$ seeds and 10.1 to $10.7\;{\mu}g/g$ seeds, respectively. 3) By washing the treated rice seeds with running water for three days, tile residual mercury concentration was reduced to 1/4 to 1/5; thus the mercury residues were 1.86 to $1.92\;{\mu}g/g$ for brown rice and 1.96 to $2.93\;{\mu}g/g$ for unhulled rice. 4) The residual mercury was present more in the unhulled rice than in the brown rice, either before or after washing of the treated seeds. 5) Among the different rice varieties, no difference was observed in mercury residues by seed treatment and washing.

• Journal of The Korean Association For Science Education
• /
• v.23 no.3
• /
• pp.286-298
• /
• 2003

The purpose of this study was to analyze scientific thinking types and processes generated in inductive inquiry by college students. Subjects were three college student. Three inductive tasks were developed: Caminalcules set I which is a task consisted of 6 imaginary animals, a potato task which is a task about the interaction between juiced potato and $H_2O_2$, and Caminalcules set 2. Subjects' thinking types and processes were investigated through thinking-aloud method and interview. Subjects' performances were recorded on videotapes and analyzed. Subjects have shown 5 types of inductive thinking in the first task; observing, discovering commonness, discovering pattern, classifying, discovering hierarchy. The processes of inductive thinking shown by students are followed; observing $\rightarrow$discovering commonness $\rightarrow$classifying $\rightarrow$discovering pattern $\rightarrow$discovering hierachy. The subtypes of inductive thinking on observing were investigated by the analysis of subjects' performance on the second task. In analysis of protocol, student' thinking types on observing have been classified as simple observing and operational observing. Operational observing has been categorized conjectural observing and predictive observing. The subtypes of inductive thinking on classification and hierarchy were investigated by the analysis of subjects' performance on the third task. In analysis of protocol, students' thinking types on classification have been searching criteria for classifying and selecting criteria for classifying. Subtypes of discovering hierarchy have been classifying groups and hierarchical ordering by students. Processes of classifying groups proceeded from searching criteria for classifying to selecting criteria for classifying.