• Korean Journal of Radiology
• /
• v.24 no.4
• /
• pp.294-304
• /
• 2023

Objective: We aimed to investigate whether image standardization using deep learning-based computed tomography (CT) image conversion would improve the performance of deep learning-based automated hepatic segmentation across various reconstruction methods. Materials and Methods: We collected contrast-enhanced dual-energy CT of the abdomen that was obtained using various reconstruction methods, including filtered back projection, iterative reconstruction, optimum contrast, and monoenergetic images with 40, 60, and 80 keV. A deep learning based image conversion algorithm was developed to standardize the CT images using 142 CT examinations (128 for training and 14 for tuning). A separate set of 43 CT examinations from 42 patients (mean age, 10.1 years) was used as the test data. A commercial software program (MEDIP PRO v2.0.0.0, MEDICALIP Co. Ltd.) based on 2D U-NET was used to create liver segmentation masks with liver volume. The original 80 keV images were used as the ground truth. We used the paired t-test to compare the segmentation performance in the Dice similarity coefficient (DSC) and difference ratio of the liver volume relative to the ground truth volume before and after image standardization. The concordance correlation coefficient (CCC) was used to assess the agreement between the segmented liver volume and ground-truth volume. Results: The original CT images showed variable and poor segmentation performances. The standardized images achieved significantly higher DSCs for liver segmentation than the original images (DSC [original, 5.40%-91.27%] vs. [standardized, 93.16%-96.74%], all P < 0.001). The difference ratio of liver volume also decreased significantly after image conversion (original, 9.84%-91.37% vs. standardized, 1.99%-4.41%). In all protocols, CCCs improved after image conversion (original, -0.006-0.964 vs. standardized, 0.990-0.998). Conclusion: Deep learning-based CT image standardization can improve the performance of automated hepatic segmentation using CT images reconstructed using various methods. Deep learning-based CT image conversion may have the potential to improve the generalizability of the segmentation network.

• Journal of Information Technology Services
• /
• v.14 no.1
• /
• pp.129-143
• /
• 2015

This study has two goals. The first goal is to investigate whether worked examples are effective in the ill-defined domain with on-line learning and the second goal is to find out which components (conceptual or procedural knowledge) of worked examples are effective factor at the given learning environment. We carried out three experiments in which Korean undergraduate or graduate students were working in three or four conditions of worked examples (CWE, PWE, CPWE, or the control group). While experiment 1 conducted in on-line learning environment did not find any effect and difference among groups and also any logical reason for those results, experiment 2 conducted in completely controlled laboratory setting with less knowledgeable students showed the clear difference among groups by the order CPWE, PWE, and CWE. Experiment 3 in which highly knowledgeable and motivated students were presented the same materials in more controlled on-line learning environment indicated the difference among groups by the order CWE, CPWE, and PWE. The results were discussed within the framework of cognitive load theory.

• Journal of Computing Science and Engineering
• /
• v.7 no.2
• /
• pp.112-121
• /
• 2013

The last years have seen a rise of interest in using electroencephalography-based brain computer interfacing methodology for investigating non-medical questions, beyond the purpose of communication and control. One of these novel applications is to examine how signal quality is being processed neurally, which is of particular interest for industry, besides providing neuroscientific insights. As for most behavioral experiments in the neurosciences, the assessment of a given stimulus by a subject is required. Based on an EEG study on speech quality of phonemes, we will first discuss the information contained in the neural correlate of this judgement. Typically, this is done by analyzing the data along behavioral responses/labels. However, participants in such complex experiments often guess at the threshold of perception. This leads to labels that are only partly correct, and oftentimes random, which is a problematic scenario for using supervised learning. Therefore, we propose a novel supervised-unsupervised learning scheme, which aims to differentiate true labels from random ones in a data-driven way. We show that this approach provides a more crisp view of the brain states that experimenters are looking for, besides discovering additional brain states to which the classical analysis is blind.

• Journal of Engineering Education Research
• /
• v.16 no.6
• /
• pp.11-18
• /
• 2013

The purpose of this study is to verify the effects of self-respect, academic self-efficacy, and career maturity on student adaptation to college and learning persistence. For this study, a web survey was conducted on the students who were in college of engineering at D college in Seoul. A total of 702 samples were analyzed for this research. The major findings of this study are as follows. First, all variables (self-respect, academic self-efficacy, career maturity) had positive effects on student adaptation to college significantly. Secondly, self-respect and academic self-efficacy had positive effects on learning persistence except career maturity. Thridly, the mediation analyses revealed that the relations between learning persistence and self-respect, academic self-efficacy, and career maturity were partially mediated by student adaptation to college. Finally, student adaptation to college had also positive effects on learning persistence. The results indicate a need to enhance student adaptation to college and design programs that support learning persistence for university students.

• Journal of The Korean Association of Information Education
• /
• v.14 no.4
• /
• pp.537-545
• /
• 2010

The purpose of this paper is to develop mathematics learning contents for elementary school 3rd graders and to verify the educational effectiveness of contents developed. An ADDIE model was applied to develop mathematics learning contents based on storytelling for concept learning. After extracting 54 concepts from the mathematics curriculum, researchers designed strategies using concepts that were combined with context which is familiar to young students. Researchers implemented a survey and interview to students and teachers to verify the effectiveness of contents. As a result, the understanding, interest, concentration, and expectation of students toward the contents developed were very high, and teachers also mentioned that these contents could be very useful teaching materials for motivation.

• Journal of Digital Contents Society
• /
• v.17 no.4
• /
• pp.265-272
• /
• 2016

Recently Chinese education market and social interest has been extended. Accordingly, smart learning based on smartphone applications became part of new educational paradigm. Also, there are more active research and development of applications for the Chinese language education. In this paper, we designed and implemented the smartphone functional application prototype for learning basic Chinese characters. Expression of Chinese characters, the comparison, listening in pronunciation, voice recording and listening, related content learning, and implement testing presented using casual user interface. In the future study, we will develop the prototype with user interface for learning Chinese conversation and individual index of evaluation can be effective learning Instrument without additional tools.

• Korean Journal of Human Ecology
• /
• v.24 no.6
• /
• pp.741-753
• /
• 2015

The purpose of this study was to investigate whether elementary children's time-use and self-regulated learning ability was different according to gender and whether children's gender and time-use effects self-regulated learning ability. Participants were 2,122 children who participated in KCYPS longitudinal study from their first grade to third grade. Time-use was reported by children's parents. Children's self-regulated learning is invented by Yang(2000). Components of self-regulated learning scale was achievement value, mastery goal orientation, action control, academic time management. The major findings were as follows. First, children's self-regulated learning was different according to chidren's gender. Girls' achievement value, mastery goal orientation, academic time management scores were higher than the boys'. Second, children's daily time was different according to their gender. Third, children's daily time-use affected their self-regulated leaning, however children's gender didn't.

• Journal of Engineering Education Research
• /
• v.18 no.1
• /
• pp.54-60
• /
• 2015

The purposes of the study are to investigate the effects on academic of the self-directed learning and in-depth interviewing program in engineering underachieved students. 17 students participated in program and the grade points average(GPA) of participants are less than 2.5. First, we focus on the change of academic achievement after the self-directed learning and in-depth interviewing program. According to results, it is very effective not only in improving academic achievement of the participation subject but also in increasing GPA. Second, the pre-survey and the post-survey were conducted to the participants. We found some facts from the difference between the pre and post surveys. The expectation and satisfaction about learning have improved after self-directed learning, and the participants' recognition showed the meaningful change in important factors about learning.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37A no.9
• /
• pp.780-785
• /
• 2012

In the current study, the authors propose a method for extracting license plate regions by means of a neural network trained to output the plates center of gravity. The method is shown to be effective. Since the learning pattern presentation positions are defined by random numbers, a different pattern is submitted to the neural network for learning each time, which enables it to form a neural network with high universality of coverage. The article discusses issues of the optimal learning surface for a license plate covered by the learning pattern, the effect of suppression learning of the number and headlight sections, as well as the effect of learning pattern enlargement/reduction and of concentration value conversion. Results of evaluation tests based on pictures of 595 vehicles taken at an underground parking garage demonstrated detection rates of 98.5%.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.7 no.4
• /
• pp.212-222
• /
• 2014

In the current study, the authors propose a method for extracting license plate regions by means of a neural network trained to output the plate's center of gravity. The method is shown to be effective. Since the learning pattern presentation positions are defined by random numbers, a different pattern is submitted to the neural network for learning each time, which enables it to form a neural network with high universality of coverage. The article discusses issues of the optimal learning surface for a license plate covered by the learning pattern, the effect of suppression learning of the number and pattern enlargement/reduction and of concentration value conversion. Results of evaluation tests based on pictures of 595 vehicles taken at an under-ground parking garage demonstrated detection rates of 98.5%, 98.7%, and 100%, respectively.