• 한국컴퓨터정보학회논문지
• /
• 제23권4호
• /
• pp.147-153
• /
• 2018

In this paper, we propose an efficient WBC 14-Diff classification which performs using the WBC-ResNet-152, a type of CNN model. The main point of view is to use Super-pixel for the segmentation of the image of WBC, and to use ResNet for the classification of WBC. A total of 136,164 blood image samples (224x224) were grouped for image segmentation, training, training verification, and final test performance analysis. Image segmentation using super-pixels have different number of images for each classes, so weighted average was applied and therefore image segmentation error was low at 7.23%. Using the training data-set for training 50 times, and using soft-max classifier, TPR average of 80.3% for the training set of 8,827 images was achieved. Based on this, using verification data-set of 21,437 images, 14-Diff classification TPR average of normal WBCs were at 93.4% and TPR average of abnormal WBCs were at 83.3%. The result and methodology of this research demonstrates the usefulness of artificial intelligence technology in the blood cell image classification field. WBC-ResNet-152 based morphology approach is shown to be meaningful and worthwhile method. And based on stored medical data, in-depth diagnosis and early detection of curable diseases is expected to improve the quality of treatment.

• Journal of Multimedia Information System
• /
• 제9권2호
• /
• pp.75-86
• /
• 2022

The existing image quality assessment (IQA) datasets have a small number of samples. Some methods based on transfer learning or data augmentation cannot make good use of image quality-related features. A No Reference (NR)-IQA method based on multi-task training and quality awareness is proposed. First, single or multiple distortion types and levels are imposed on the original image, and different strategies are used to augment different types of distortion datasets. With the idea of weak supervision, we use the Full Reference (FR)-IQA methods to obtain the pseudo-score label of the generated image. Then, we combine the classification information of the distortion type, level, and the information of the image quality score. The ResNet50 network is trained in the pre-train stage on the augmented dataset to obtain more quality-aware pre-training weights. Finally, the fine-tuning stage training is performed on the target IQA dataset using the quality-aware weights to predicate the final prediction score. Various experiments designed on the synthetic distortions and authentic distortions datasets (LIVE, CSIQ, TID2013, LIVEC, KonIQ-10K) prove that the proposed method can utilize the image quality-related features better than the method using only single-task training. The extracted quality-aware features improve the accuracy of the model.

• The Journal of Asian Finance, Economics and Business
• /
• 제9권4호
• /
• pp.143-152
• /
• 2022

The purpose of this paper is to investigate the impact of several factors on students' decisions to attend a public institution in Vietnam's North Central area. The enrollment issue toward regional institutions is particularly critical in the Vietnam Ministry of Education and Training reforming the university enrollment process and the complicated scenario of the Covid-19 pandemic. A total of 500 students were surveyed for research samples. Data is synthesized, validated, cleaned, and analyzed using SPSS and AMOS software using methods including reliability, EFA, CFA, and SEM. The findings suggest that the proposed independent components (individual factors, study fees, advertisement, infrastructure and facilities, local features, and lastly, training activities) have a beneficial impact on students' decision to attend a public university in the North Central region. The study also found that the graduation exam outcome had a moderating effect on the relationship between registration and students' decisions. These imply targeted solutions for regional public universities to diversify training majors, improve training quality, capitalize on local advantages, increase interaction, and promote training programs and image to be more effective in attracting students and maintaining competition in the current enrollment environment.

• 대한물리의학회지
• /
• 제17권1호
• /
• pp.31-39
• /
• 2022

PURPOSE: This study examined the changes in the blood fatigue variables caused by isokinetic and isotonic exercise training. METHODS: Ten healthy adult males with at least one year of athletic experience participated. The participants performed the isokinetic circuit exercise program first, followed by an isotonic circuit exercise program. A two-hour break was allowed between the isokinetic circuit exercise program and the isotonic circuit exercise program. The circuit exercise program consisted of four items (Squat, Deadlift, Shoulder press, and Bench press). The blood samples were analyzed for the LDH, CPK, and Cortisol levels. RESULTS: The LDH level in the isokinetic group was significantly different from the isotonic group. In particular, the change in LDH level in the isokinetic group was 33.30% lower than that of the isotonic group. The serum CPK level of the isokinetic group showed a 10.03% lower decrease than the isotonic group, but the difference was not significant. The Cortisol level was relatively unchanged in the isotonic group, but it decreased in the isokinetic group. On the other hand, the Cortisol level did not show a significant difference between the two groups. CONCLUSION: The isokinetic group showed alleviation of the three indices, unlike the isotonic group. Further studies associated with the changes in blood fatigue variables through various exercise programs and exercise intensity will be needed.

• Smart Structures and Systems
• /
• 제33권1호
• /
• pp.65-91
• /
• 2024

The use of five optimization techniques for the prediction of a strength-based concrete mixture's best-fit model is examined in this work. Five optimization techniques are utilized for this purpose: Slime Mold Algorithm (SMA), Black Hole Algorithm (BHA), Multi-Verse Optimizer (MVO), Vortex Search (VS), and Whale Optimization Algorithm (WOA). MATLAB employs a hybrid learning strategy to train an artificial neural network that combines least square estimation with backpropagation. Thus, 72 samples are utilized as training datasets and 31 as testing datasets, totaling 103. The multi-layer perceptron (MLP) is used to analyze all data, and results are verified by comparison. For training datasets in the best-fit models of SMA-MLP, BHA-MLP, MVO-MLP, VS-MLP, and WOA-MLP, the statistical indices of coefficient of determination (R²) in training phase are 0.9603, 0.9679, 0.9827, 0.9841 and 0.9770, and in testing phase are 0.9567, 0.9552, 0.9594, 0.9888 and 0.9695 respectively. In addition, the best-fit structures for training for SMA, BHA, MVO, VS, and WOA (all combined with multilayer perceptron, MLP) are achieved when the term population size was modified to 450, 500, 250, 150, and 500, respectively. Among all the suggested options, VS could offer a stronger prediction network for training MLP.

• 한국보건간호학회지
• /
• 제26권3호
• /
• pp.417-427
• /
• 2012

Purpose: The purpose of this study was to examine the effect of a 12-week combined exercise training program on the body composition, physical fitness levels, and metabolic syndrome profiles of obese women. Methods: Twelve obese women were assigned to the combined exercise training program group. The women underwent training for 70-90 min/d, three times per week for a period of 12 weeks. Paired samples t-tests were performed using SPSS ver. 17.0 for analysis of the results. Results: The results of this study showed that body-composition parameters such as weight, fat-free mass, body fat mass, body-mass index, body fat, waist-hip ratio, basal metabolic rate, and intra-abdominal fat, physical fitness parameters such as muscle strength, muscle endurance, flexibility, and cardiac endurance, and metabolic syndrome biomarkers such as triglyceride levels, high-density lipoprotein cholesterol levels, glucose levels, systolic blood pressure, and waist circumference before participation the training program differed significantly from those after participation in the training program (p<0.05). However, diastolic blood pressure before participation in the training program did not differ significantly from that after participation in the training program (p>0.05). Conclusion: We concluded that a 12-week combined exercise training program could be a good exercise program for improvement of the body composition, physical fitness levels, and metabolic syndrome profiles of obese women.

• Journal of Gastric Cancer
• /
• 제19권3호
• /
• pp.301-314
• /
• 2019

Purpose: Peritoneal carcinomatosis in gastric cancer (GC) patients results in extremely poor prognosis. Malignant ascites samples are the most appropriate biological material to use to evaluate biomarkers for peritoneal carcinomatosis. This study identified exosomal MicroRNAs (miRNAs) differently expressed between benign liver cirrhosis-associated ascites (LC-ascites) and malignant gastric cancer-associated ascites (GC-ascites), and validated their role as diagnostic biomarkers for GC-ascites. Materials and Methods: Total RNA was extracted from exosomes isolated from 165 ascites samples (73 LC-ascites and 92 GC-ascites). Initially, microarrays were used to screen the expression levels of 2,006 miRNAs in the discovery cohort (n=22). Subsequently, quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) analyses were performed to validate the expression levels of selected exosomal miRNAs in the training (n=70) and validation (n=73) cohorts. Furthermore, carcinoembryonic antigen (CEA) levels were determined in ascites samples. Results: The miR-574-3p, miR-181b-5p, miR-4481, and miR-181d were significantly downregulated in the GC-ascites samples compared to the LC-ascites samples, and miR-181b-5p showed the best diagnostic performance for GC-ascites (area under the curve [AUC]=0.798 and 0.846 for the training and validation cohorts, respectively). The diagnostic performance of CEA for GC-ascites was improved by the combined analysis of miR-181b-5p and CEA (AUC=0.981 and 0.946 for the training and validation cohorts, respectively). Conclusions: We identified exosomal miRNAs capable of distinguishing between non-malignant and GC-ascites, showing that the combined use of miR-181b-5p and CEA could improve diagnosis.

• 전기학회논문지
• /
• 제60권1호
• /
• pp.205-213
• /
• 2011

The Long Term Evolution (LTE) system is designed to provide a high quality data service for fast moving mobile users. It is based on the Orthogonal Frequency Division Multiplexing (OFDM) and relies its channel estimation on the training samples which are systematically built within the transmitting data. Either a preamble or a lattice type is used for the distribution of training samples and the latter suits better for the multipath fading channel environment whose channel frequency response (CFR) fluctuates rapidly with time. In the lattice-type structure, the estimation of the CFR makes use of the least squares estimate (LSE) for each pilot samples, followed by an interpolation both in time-and in frequency-domain to fill up the channel estimates for subcarriers corresponding to data samples. All interpolation schemes should rely on the pilot estimates only, and thus, their performances are bounded by the quality of pilot estimates. However, the additive noise give rise to high fluctuation on the pilot estimates, especially in a communication environment with low signal-to-noise ratio. These high fluctuations could be monitored in the alternating high values of the first forward differences (FFD) between pilot estimates. In this paper, we analyzed statistically those FFD values and propose a postprocessing algorithm to suppress high fluctuations in the noisy pilot estimates. The proposed method is based on a localized adaptive moving-average filtering. The performance of the proposed technique is verified on a multipath environment suggested on a 3GPP LTE specification. It is shown that the mean-squared error (MSE) between the actual CFR and pilot estimates could be reduced up to 68% from the noisy pilot estimates.

• Structural Engineering and Mechanics
• /
• 제83권3호
• /
• pp.293-304
• /
• 2022

In this work, a novel reliability approach for combined high and low cycle fatigue (CCF) estimation is developed by combining active learning strategy with least squares support vector machines (LS-SVM) (named as ALS-SVM) surrogate model to address the multi-resources uncertainties, including working loads, material properties and model itself. Initially, a new active learner function combining LS-SVM approach with Monte Carlo simulation (MCS) is presented to improve computational efficiency with fewer calls to the performance function. To consider the uncertainty of surrogate model at candidate sample points, the learning function employs k-fold cross validation method and introduces the predicted variance to sequentially select sampling. Following that, low cycle fatigue (LCF) loads and high cycle fatigue (HCF) loads are firstly estimated based on the training samples extracted from finite element (FE) simulations, and their simulated responses together with the sample points of model parameters in Coffin-Manson formula are selected as the MC samples to establish ALS-SVM model. In this analysis, the MC samples are substituted to predict the CCF reliability of turbine blades by using the built ALS-SVM model. Through the comparison of the two approaches, it is indicated that the reliability model by linear cumulative damage rule provides a non-conservative result compared with that by the proposed one. In addition, the results demonstrate that ALS-SVM is an effective analysis method holding high computational efficiency with small training samples to gain accurate fatigue reliability.

• International Journal of Industrial Entomology and Biomaterials
• /
• 제11권1호
• /
• pp.63-69
• /
• 2005

While observing silkworm larval samples received from field, microsporidian spores formed within the haemocytes of silkworm haemolymph were observed. The spores of microsporidian sp. were purified and characterized for morphological characters viz., size, shape as well as serological affinity with different Nosema spp. (M$_{11}$ and M$_{12}$). The infectivity of the isolated spores to silkworm was also studied. The microsporidian sp. was found to be highly pathogenic to silkworm, B. mori. The isolated microsporidian sp. was designated as NIK-5hm, which formed ovocylindrical spore in the haemocytes of silkworm and differed in spore size (length, 4.55 $\mu$m & width, 2.10 $\mu$m) and shape from Nosema bombycis (NIK-ls), NIK-2r (Nosema sp. Mysore [3.6 & 2.8 $\mu$m]), NIK-3h (Nosema sp. M$_{11}$ [3.8 & 1.8 $\mu$m]), NIK-4m (Nosema sp. M$_{12}$ [5.0 & 2.1 $\mu$m]) and Lb$_{ms}$ (Nosema sp. in Lamerine breed of silkworm [4.36 & 2.14]). In immonological test (Latex agglutination test), the isolated microsporidian spores did not react with antibody sensitized latex particles of N. bombycis, M$_{11}$, M$_{12}$ and Lb$_{ms}$ and thus are different type of microsporidian sp., parasitic to silkworm, Bombyx mori L.