• Journal of Multimedia Information System
• /
• v.8 no.3
• /
• pp.175-182
• /
• 2021

Oral cancer is ranked second most diagnosed cancer among Indian population and ranked sixth all around the world. Oral cancer is one of the deadliest cancers with high mortality rate and very less 5-year survival rates even after treatment. It becomes necessary to detect oral malignancies as early as possible so that timely treatment may be given to patient and increase the survival chances. In recent years deep learning based frameworks have been proposed by many researchers that can detect malignancies from medical images. In this paper we have proposed a deep learning-based framework which detects oral cancer from histopathology images very efficiently. We have designed our model to split the color channels and extract deep features from these individual channels rather than single combined channel with the help of Efficient NET B3. These features from different channels are fused by using feature fusion module designed as a layer and placed before dense layers of Efficient NET. The experiments were performed on our own dataset collected from hospitals. We also performed experiments of BreakHis, and ICML datasets to evaluate our model. The results produced by our model are very good as compared to previously reported results.

• ETRI Journal
• /
• v.44 no.3
• /
• pp.476-490
• /
• 2022

With the recent advances in technology, automatic speech recognition (ASR) has been widely used in real-world applications. The efficiency of converting large amounts of speech into text accurately with limited resources has become more vital than ever. In this study, we propose a method to rapidly recognize a large speech database via a transformer-based end-to-end model. Transformers have improved the state-of-the-art performance in many fields. However, they are not easy to use for long sequences. In this study, various techniques to accelerate the recognition of real-world speeches are proposed and tested, including decoding via multiple-utterance-batched beam search, detecting end of speech based on a connectionist temporal classification (CTC), restricting the CTC-prefix score, and splitting long speeches into short segments. Experiments are conducted with the Librispeech dataset and the real-world Korean ASR tasks to verify the proposed methods. From the experiments, the proposed system can convert 8 h of speeches spoken at real-world meetings into text in less than 3 min with a 10.73% character error rate, which is 27.1% relatively lower than that of conventional systems.

• Archives of Craniofacial Surgery
• /
• v.24 no.4
• /
• pp.145-158
• /
• 2023

Vascular anomalies encompass a variety of malformations and tumors that can result in severe morbidity and mortality in both adults and children. Advances have been made in the classification and diagnosis of these anomalies, with the International Society for the Study of Vascular Anomalies establishing a widely recognized classification system. In recent years, notable progress has been made in genetic testing and imaging techniques, enhancing our ability to diagnose these conditions. The increasing sophistication of genetic testing has facilitated the identification of specific genetic mutations that help treatment decisions. Furthermore, imaging techniques such as magnetic resonance imaging and computed tomography have greatly improved our capacity to visualize and detect vascular abnormalities, enabling more accurate diagnoses. When considering reconstructive surgery for facial vascular anomalies, it is important to consider both functional and cosmetic results of the procedure. Therefore, a comprehensive multidisciplinary approach involving specialists from dermatology, radiology, and genetics is often required to ensure effective management of these conditions. Overall, the treatment approach for facial vascular anomalies depends on the type, size, location, and severity of the anomaly. A thorough evaluation by a team of specialists can determine the most appropriate and effective treatment plan.

• Nuclear Engineering and Technology
• /
• v.56 no.2
• /
• pp.558-567
• /
• 2024

Diagnostic support systems are being researched to assist operators in identifying and responding to abnormal events in a nuclear power plant. Most studies to date have considered single abnormal events only, for which it is relatively straightforward to obtain data to train the deep learning model of the diagnostic support system. However, cases in which multiple abnormal events occur must also be considered, for which obtaining training data becomes difficult due to the large number of combinations of possible abnormal events. This study proposes an approach to maintain diagnostic performance for multiple abnormal events by training a deep learning model with data on single abnormal events only. The proposed approach is applied to an existing algorithm that can perform feature selection and multi-label classification. We choose an extremely randomized trees classifier to select dedicated monitoring parameters for target abnormal events. In diagnosing each event occurrence independently, two-channel convolutional neural networks are employed as sub-models. The algorithm was tested in a case study with various scenarios, including single and multiple abnormal events. Results demonstrated that the proposed approach maintained diagnostic performance for 15 single abnormal events and significantly improved performance for 105 multiple abnormal events compared to the base model.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.10
• /
• pp.43-57
• /
• 2022

In this paper, we propose a method of analyzing research papers published by researchers belonging to university departments in the field of disaster & safety for the scientometric analysis of the research status in the field of disaster safety. In order to conduct analysis research, the dataset constructed in previous studies was newly improved and utilized. In detail, for research papers of authors belonging to the disaster prevention and safety engineering type department of domestic universities, institution identification, cited journal identification of references, department type classification, disaster safety type classification, researcher major information, KSIC(Korean Standard Industrial Classification) mapping information was reflected in the experimental data. The proposed method has a difference from previous studies in the field of disaster & safety and data set based on related keyword searches. As a result of the analysis, the type and regional distribution of organizations belonging to the department of disaster prevention and safety engineering, the composition of co-authored department types, the researchers' majors, the status of disaster safety types and standard industry classification, the status of citations in academic journals, and major keywords were identified in detail. In addition, various co-occurrence networks were created and visualized for each analysis unit to identify key connections. The research results will be used to identify and recommend major organizations and information by disaster type for the establishment of an intelligent crisis warning system. In order to provide comprehensive and constant analysis information in the future, it is necessary to expand the analysis scope and automate the identification and classification process for data set construction.

• PNF and Movement
• /
• v.14 no.3
• /
• pp.219-229
• /
• 2016

Purpose: Patients with Parkinson's disease suffer many restrictions in daily life. This case report investigated how intervention with proprioceptive neuromuscular facilitation (PNF)-based physical therapy can affect the activities of Parkinson's patients. Methods: The subject was a 67-year-old female patient diagnosed with Parkinson's disease 3 years ago. Since the last five months, overall activities have become difficult for her, and she finally visited a hospital to take outpatient physical therapy because of the gait difficulties she suffered. The patient's medical history, system review, body structure and function, and activities were evaluated. The patient had difficulties in activities such as lying down, sitting, standing, maintaining a standing position, and walking. The PNF-based intervention was used for treating the impairments and improving the activities. The intervention was performed for 30 min a day, three times a week, for eight weeks. The qualifier of the international classification of functioning, disability, and health was used to measure the result, and the measurement was conducted before and after the intervention for eight weeks. Results: According to the result, the scores for maintaining a standing position, moving around within the home, and going to the toilet improved to "no problem" from "moderate problem." The scores for shifting the body's center of gravity, walking short distances, and washing oneself improved from "moderate problem" to "mild problem." The scores for sitting and standing improved to "no problem" and "mild problem" from "complete problem." The scores for preparing meals and doing housework improved from "severe problems" to "mild problem" or "moderate problem." The scores for walking long distances, moving around outside the home and other buildings, and using transportation did not show significant changes. Conclusion: Intervention with PNF-based physical therapy improved the activity of patients with Parkinson's disease, thus proving its effectiveness. The case report suggested that a therapist can use PNF as a physical therapy intervention for patients with Parkinson's who suffered restrictions in daily activities.

• Journal of the Korean Society of Physical Medicine
• /
• v.18 no.1
• /
• pp.77-86
• /
• 2023

PURPOSE: This study developed an experimental algorithm, which is similar or identical to semantic linking for KCF codes, even if it converted existing semantic code linking methods to morphological code extraction methods. The purpose of this study was to verify the applicability of the system. METHODS: An experimental algorithm was developed as a morphological extraction method using code-specific words in the KCF code descriptions. The algorithm was designed in five stages that extracted KCF code using natural language paragraphs. For verification, 80 clinical natural language experimental cases were defined. Data acquisition for the study was conducted with the deliberation and approval of the bioethics committee of the relevant institution. Each case was linked by experts and was extracted through the System. The linking accuracy index model was used to compare the KCF code linking by experts with those extracted from the system. RESULTS: The accuracy was checked using the linking accuracy index model for each case. The analysis was divided into five sections using the accuracy range. The section with less than 25% was compared; the first experimental accuracy was 61.24%. In the second, the accuracy was 42.50%. The accuracy was improved to 30.59% in the section by only a weight adjustment. The accuracy can be improved by adjusting several independent variables applied to the system. CONCLUSION: This paper suggested and verified a way to easily extract and utilize KCF codes even if they are not experts. KCF requires the system for utilization, and additional study will be needed.

• Journal of Digital Convergence
• /
• v.12 no.12
• /
• pp.365-370
• /
• 2014

The existing system consumes a considerable amount time and cost for extracting the patterns from whole sequences or misaligned sequences. In this paper, We propose the classification system, which creates the partition sequence sections using multiple sequence alignment method and extracts the sequential patterns from these section. These extracted patterns are accumulated motif candidate sets and then used the training sets of SVM classifier. This proposed system predicts a HPV-type(high/low) using the learned knowledges from known/unknown protein sequences and shows more improved precision, recall than previous system in 30% minimum support.

• Journal of Korean Society for Atmospheric Environment
• /
• v.25 no.3
• /
• pp.196-210
• /
• 2009

The land surface precesses is very important to predict urban meteorological conditions. Thus, the latest land use data set to reflect the rapid progress in urbanization was applied to simulate urban thermal environment in Daegu. Because use of the U.S geological Survey (USGS) 25-category data, currently in the fifth-generation Pennsylvania State University-National Center for Atmospheric Research Mesoscale Model (MM5), does not accurately described the heterogeneity of urban surface, we replaced the land use data in USGS with the latest land-use data of the Korea Ministry of Environment over Daegu. The single urban category in existing 24-category U.S. Geological survey land cover classification used in MM5 was divided into 5 classes to account for heterogeneity of urban land cover. The new land cover classification (MC-LULC) improved the capability of MM5 to simulate the daytime part of the diurnal temperature cycle in the urban area. The 'MC-LULC' simulation produced the observed temperature field reasonably well, including spatial characteristics. The warm cores in western Daegu is characterized by an industrial area.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.65 no.1
• /
• pp.15-26
• /
• 2023

This study aimed to evaluate the performance of water level classification from CCTV images in agricultural facilities such as reservoirs. Recently, the CCTV system, widely used for facility monitor or disaster detection, can automatically detect and identify people and objects from the images by developing new technologies such as a deep learning system. Accordingly, we applied the ResNet-50 deep learning system based on Convolutional Neural Network and analyzed the water level of the agricultural reservoir from CCTV images obtained from TOMS (Total Operation Management System) of the Korea Rural Community Corporation. As a result, the accuracy of water level detection was improved by excluding night and rainfall CCTV images and applying measures. For example, the error rate significantly decreased from 24.39 % to 1.43 % in the Bakseok reservoir. We believe that the utilization of CCTVs should be further improved when calculating the amount of water supply and establishing a supply plan according to the integrated water management policy.