• Journal of the Korean Society of Radiology
• /
• v.14 no.6
• /
• pp.773-780
• /
• 2020

With the development of the 4th industrial, research is being conducted to prevent diseases and reduce damage in various fields of science and technology such as medicine, health, and bio. As a result, artificial intelligence technology has been introduced and researched for image analysis of radiological examinations. In this paper, we will directly apply a deep learning model for classification and detection of pneumonia using chest X-ray images, and evaluate whether the deep learning model of the Inception series is a useful model for detecting pneumonia. As the experimental material, a chest X-ray image data set provided and shared free of charge by Kaggle was used, and out of the total 3,470 chest X-ray image data, it was classified into 1,870 training data sets, 1,100 validation data sets, and 500 test data sets. I did. As a result of the experiment, the result of metric evaluation of the Inception V3 deep learning model was 94.80% for accuracy, 97.24% for precision, 94.00% for recall, and 95.59 for F1 score. In addition, the accuracy of the final epoch for Inception V3 deep learning modeling was 94.91% for learning modeling and 89.68% for verification modeling for pneumonia detection and classification of chest X-ray images. For the evaluation of the loss function value, the learning modeling was 1.127% and the validation modeling was 4.603%. As a result, it was evaluated that the Inception V3 deep learning model is a very excellent deep learning model in extracting and classifying features of chest image data, and its learning state is also very good. As a result of matrix accuracy evaluation for test modeling, the accuracy of 96% for normal chest X-ray image data and 97% for pneumonia chest X-ray image data was proven. The deep learning model of the Inception series is considered to be a useful deep learning model for classification of chest diseases, and it is expected that it can also play an auxiliary role of human resources, so it is considered that it will be a solution to the problem of insufficient medical personnel. In the future, this study is expected to be presented as basic data for similar studies in the case of similar studies on the diagnosis of pneumonia using deep learning.

• Journal of Korean Medicine for Obesity Research
• /
• v.13 no.2
• /
• pp.58-65
• /
• 2013

Objectives: The objective of this study was to analyse the methods being used to control food intake and physical activity in RCTs of human obesity. Methods: A total of 21 randomized controlled trials (RCTs) were investigated. Nine of which were domestic studies from "http://oasis.kiom.re.kr" and the other of which were foreign studies from systematic reviews of RCTs on herbal medicine for treatment of human obesity. Results: According to domestic studies, "low calorie diet" were recommended in five cases of the domestic studies, "maintain current dietary habit" were recommended in two and no information on diet was two. Considering the seven cases where the information on diet was available, patients' food intake were checked at every visit in six cases. Only two cases among the six had been dropped owing to the violation of dietary habit by patients. Exercises were prohibited in two cases, "maintain current level of phisical activity" were recommended in three cases and, from the rest, no information was available. The level of physical activity were not strictly controlled by any means hence no drop out. According to foreign studies, "low calorie diet" were recommended in two cases, "very low calorie diet (less than 700 kcal/day)" in one case, "maintain current dietary habit" in two cases, "do not eat fat" in two cases and no information was available in the rest five cases. Exercises which concerns spending about 300 kcal/day was recommended in one case, "moderate exercise" were recommended in three cases, "maintain current level of physical activity" were recommended in three cases and no information available in the rest five cases. Conclusions: In order to improve the accuracy of RCT, for the dietary side, researchers should record patient food intake at every visit by means of 24-hour dietary recall methods. This can be supplemented by multiple choice survey that are designed to help patients to diagnose themselves more accurately leading to less bias. For the exercise side, it is highly recommended to confine the exercises to walking only so as to quantify the amount of physical activity more easily by using pedometer.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.30 no.4
• /
• pp.617-629
• /
• 2020

In order to overcome the limitations of the rule-based intrusion detection system due to changes in Internet computing environments, the emergence of new services, and creativity of attackers, network anomaly detection (NAD) using machine learning and deep learning technologies has received much attention. Most of these existing machine learning and deep learning technologies for NAD use supervised learning methods to learn a set of training data set labeled 'normal' and 'attack'. This paper presents the feasibility of the unsupervised learning AutoEncoder(AE) to NAD from data sets collecting of secured network traffic without labeled responses. To verify the performance of the proposed AE mode, we present the experimental results in terms of accuracy, precision, recall, f1-score, and ROC AUC value on the NSL-KDD training and test data sets. In particular, we model a reference AE through the deep analysis of diverse AEs varying hyper-parameters such as the number of layers as well as considering the regularization and denoising effects. The reference model shows the f1-scores 90.4% and 89% of binary classification on the KDDTest+ and KDDTest-21 test data sets based on the threshold of the 82-th percentile of the AE reconstruction error of the training data set.

• Journal of Community Nutrition
• /
• v.5 no.2
• /
• pp.105-111
• /
• 2003

Balanced intake of appropriate nutrients is the key to sustaining and promoting health as well as preventing and treating diseases. It is not always easy, however, to take balanced nutrition and various related factors must be taken into consideration. This report provides a summary of studies that have examined the nutrient intakes of Japanese middle-aged men and the relationship of this intake to drinking, job-related separation from families, and health practices. The alcohol consumption of Japanese middle-aged men has more than tripled in the last forty years. When nutrient intake was examined in relation to alcohol consumption, it was discovered that the maximum acceptable alcohol consumption was approximately 23 grams (about two drinks) of pure alcohol, provided the level of nutrient intake for drinkers was equal to that of non-drinkers. The alcohol energy ratio was approximately 5％. It was also discovered that middle-aged men's eating habits deteriorate when they relocate to new posts without their families and live by themselves. Compared to those living with their families, a higher proportion of those living alone have unfavorable eating habits including skipping breakfast or lunch, having a late lunch, and eating and drinking after dinner until bedtime. When Breslow's seven health practices, nutrient intake, and consumption weight by food group were examined, it was discovered that the group that had many beneficial eating and living habits consumed plenty of legume, pulses, fruit, green yellow vegetables and milk products. Their intake of vitamins and minerals was high and the results of a physical examination proved to be excellent. According to nutrition surveys conducted in Japan, China, the United Kingdom and the United States using a 24-hour recall method with common protocols and strict controls to ensure high levels of accuracy and cross-study validity, the Japanese had the highest cholesterol intake and the lowest dietary fiber intake among the four countries. Also, the alcohol energy ratio of the Japanese exceeded 8％, the highest among the four countries, while their intake of magnesium and iron was the lowest These results indicate that it is necessary to enhance nutritional education for middle-aged men and to reinforce the social environments in which they live and work in order to promote proper diet and nutrition in Japan. (J Community Nutrition 5(2) : 105-111, 2003)

• Journal of Nutrition and Health
• /
• v.43 no.6
• /
• pp.638-652
• /
• 2010

A valid food or dish frequency questionnaire needs to be developed in Korea for accurate dietary assessment because the dietary practices of Koreans are very different from those of other countries. This study was conducted to evaluate the accuracy of the newly developed, semi-quantitative, dish-based frequency questionnaire (Semi-DFQ) with 12-day dietary records (12-DRs) as a gold standard. The study subjects were 115 men and 173 women aged 30-65 years old. We calibrated the frequency, portion size and daily intake of 112 dish items reported in Semi-DFQ with those in 12-DRs by Spearman rank correlation coefficients (SCCs). The consumption frequency and portion size reported in Semi-DFQ were higher than those in 12-DRs. The SCCs for the consumption frequency of various dishes ranged from -0.07 (fried seaweed) to 0.70 (instant coffee), the portion size ranged from -0.09 (cold seaweed soup) to 0.68 (soju), and the daily intake ranged from -0.07 (fried seaweed) to 0.71 (soju). The SCCs were higher for dishes consumed daily, such as steamed rice,milk, coffee and alcohol, than those of foods eaten rarely. The overall agreements between the Semi-DFQ and 12-DRs were low for categories of consumption frequency and portion size, even though some dishes showed high SCCs. The SCCs of the two methods in consumption frequency and amount were higher among the women and younger subject. The results revealed the limitation of the Semi-DFQ for evaluating the status of usual individual intake. Therefore, the Semi-DFQ can be used in addition as dietary records and 24-hour recall depending on the research aims.

• Journal of Korea Multimedia Society
• /
• v.13 no.7
• /
• pp.1044-1050
• /
• 2010

Endoscopic ultrasonography is a medical procedure in endoscopy combined with ultrasound to obtain images of the internal organs. It is useful to have a predictive pathological manifestation since a doctor can observe tumors under mucosa. However, it is often subjective to judge the degree of malignant degeneration of tumors. Thus, in this paper, we propose a feature analysis procedure to make the pathological manifestation more objective so as to improve the accuracy and recall of the diagnosis. In the process, we extract the ultrasound region from the image obtained by endoscopic ultrasonography. It is necessary to standardize the intensity of this region with the intensity of water region as a base since frequently found small intensity difference is only to be inefficient in the analysis. Then, we analyze the spot region with high echo and calcium deposited region by applying LVQ algorithm and bit plane partitioning procedure to tumor regions selected by medical expert. For detailed analysis, features such as intensity value, intensity information included within two random points chosen by medical expert in tumor region, and the slant of outline of tumor region in order to decide the degree of malignant degeneration. Such procedure is proven to be helpful for medical experts in tumor analysis.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.390-397
• /
• 2009

Endoscopic ultrasonography is a medical procedure in endoscopy combined with ultrasound to obtain images of the internal organs. It is useful to have a predictive pathological manifestation since a doctor can observe tumors under mucosa. However, it is often subjective to judge the degree of malignant degeneration of tumors. Thus, in this paper, we propose a feature analysis procedure to make the pathological manifestation more objective so as to improve the accuracy and recall of the diagnosis. In the process, we extract the ultrasound region from the image obtained by endoscopic ultrasonography. It is necessary to standardize the intensity of this region with the intensity of water region as a base since frequently found small intensity difference is only to be inefficient in the analysis. Then, we analyze the spot region with high echo and calcium deposited region by applying LVQ algorithm and bit plane partitioning procedure to tumor regions selected by medical expert. For detailed analysis, features such as intensity value, intensity information included within two random points chosen by medical expert in tumor region, and the slant of outline of tumor region in order to decide the degree of malignant degeneration. Such procedure is proven to be helpful for medical experts in tumor analysis.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.12
• /
• pp.491-498
• /
• 2019

Digital twin is a technology that virtualizes physical objects of the real world on a computer. It is used by collecting sensor data through IoT, and using the collected data to connect physical objects and virtual objects in both directions. It has an advantage of minimizing risk by tuning an operation of virtual model through simulation and responding to varying environment by exploiting experiments in advance. Recently, artificial intelligence and machine learning technologies have been attracting attention, so that tendency to virtualize a behavior of physical objects, observe virtual models, and apply various scenarios is increasing. In particular, recognition of each robot's motion is needed to build digital twin for co-robot which is a heart of industry 4.0 factory automation. Compared with modeling based research for recognizing motion of co-robot, there are few attempts to predict motion based on sensor data. Therefore, in this paper, an experimental environment for collecting current and inertia data in co-robot to detect the motion of the robot is built, and a motion prediction model based on the collected sensor data is proposed. The proposed method classifies the co-robot's motion commands into 9 types based on joint position and uses current and inertial sensor values to predict them by accumulated learning. The data used for accumulating learning is the sensor values that are collected when the co-robot operates with margin in input parameters of the motion commands. Through this, the model is constructed to predict not only the nine movements along the same path but also the movements along the similar path. As a result of learning using SVM, the accuracy, precision, and recall factors of the model were evaluated as 97% on average.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.8
• /
• pp.187-195
• /
• 2013

Most of former studies for car license plate detection restrict the image acquisition environment. The aim of this research is to diminish the restrictions by proposing a new method of using SIFT and neural network. SIFT can be used in diverse situations with less restriction because it provides size- and rotation-invariance and large discriminating power. SIFT extracted from the license plate image is divided into the internal(inside class) and the external(outside class) ones and the classifier is trained using them. In the proposed method, by just putting the various types of license plates, the trained neural network classifier can process all of the types. Although the classification performance is not high, the inside class appears densely over the plate region and sparsely over the non-plate regions. These characteristics create a local feature map, from which we can identify the location with the global maximum value as a candidate of license plate region. We collected image database with much less restriction than the conventional researches. The experiment and evaluation were done using this database. In terms of classification accuracy of SIFT keypoints, the correct recognition rate was 97.1%. The precision rate was 62.0% and recall rate was 50.2%. In terms of license plate detection rate, the correct recognition rate was 98.6%.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.38 no.4
• /
• pp.353-361
• /
• 2020

Due to the rapid urbanization, various traffic problems such as traffic jams during commute and regular traffic jams are occurring. In order to solve these traffic problems, it is necessary to quickly and accurately estimate and analyze traffic volume. ITS (Intelligent Transportation System) is a system that performs optimal traffic management by utilizing the latest ICT (Information and Communications Technology) technologies, and research has been conducted to analyze fast and accurate traffic volume through various techniques. In this study, we proposed a deep learning-based vehicle detection method using UAV (Unmanned Aerial Vehicle) video for real-time traffic analysis with high accuracy. The UAV was used to photograph orthogonal videos necessary for training and verification at intersections where various vehicles pass and trained vehicles by classifying them into sedan, truck, and bus. The experiment on UAV dataset was carried out using YOLOv3 (You Only Look Once V3), a deep learning-based object detection technique, and the experiments achieved the overall object detection rate of 90.21%, precision of 95.10% and the recall of 85.79%.