• The Korean Journal of Nuclear Medicine
• /
• v.38 no.5
• /
• pp.338-343
• /
• 2004

Purpose: The purpose of this study was to develop a small animal PET using dual layer phoswich detector to minimize parallax error that degrades spatial resolution at the outer part of field-of-view (FOV). Materials and Methods: A simulation tool GATE (Geant4 Application for Tomographic Emission) was used to derive optimal parameters of small PET, and PET was developed employing the parameters. Lutetium Oxyorthosilicate (LSO) and Lutetium-Yttrium Aluminate-Perovskite(LuYAP) was used to construct dual layer phoswitch crystal. $8{\times}8$ arrays of LSO and LuYAP pixels, $2mm{\times}2mm{\times}8mm$ in size, were coupled to a 64-channel position sensitive photomultiplier tube. The system consisted of 16 detector modules arranged to one ring configuration (ring inner diameter 10 cm, FOV of 8 cm). The data from phoswich detector modules were fed into an ADC board in the data acquisition and preprocessing PC via sockets, decoder block, FPGA board, and bus board. These were linked to the master PC that stored the events data on hard disk. Results: In a preliminary test of the system, reconstructed images were obtained by using a pair of detectors and sensitivity and spatial resolution were measured. Spatial resolution was 2.3 mm FWHM and sensitivity was 10.9 $cps/{\mu}Ci$ at the center of FOV. Conclusion: The radioactivity distribution patterns were accurately represented in sinograms and images obtained by PET with a pair of detectors. These preliminary results indicate that it is promising to develop a high performance small animal PET.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.28 no.3
• /
• pp.471-479
• /
• 2001

The purpose of this study was to compare the effect of the high intensity halogen light $(850\sim1000mW/cm^2)$ with that of the conventional halogen light $(400mW/cm^2)$ on the hardness of composite resin. Three resin composites (Z-100, 3M, U.S.A. : Tetric Ceram, Vivadent, Liechtenstein; SureFil, Dentsply, U.S.A.) were filed in the stainless steel moulds which were 4mm in diameter and 2, 3, 4, and 5mm in depth, respectively. They were cured under the four different modes : (1) conventional mode, 40 seconds at $400mW/cm^2$; (2) 'ramp' mode, 10 seconds at 100 to $1000mW/cm^2$ plus 10 seconds at $1000mW/cm^2$; (3) 'boost' mode, 10 seconds at $1000mW/cm^2$; and (4) 'standard' mode, 20 seconds at $850mW/cm^2$. The surface hardnesses of the top and the bottom of the resin samples were measured with a microhardness tester (MXT70, Matsuzawa, Japan). The top surface hardness was not significantly different among the curing modes. The bottom surface hardness was generally the highest in the conventional mode and the lowest in the high intensity boost mode. There was no significant difference in the bottom surface hardness between the conventional mode and the high intensity standard mode in 2mm depth. The results suggest that the curing time of the high intensity halogen light $(850mW/cm^2)$ should be at least 20 seconds to produce the equal level of the bottom surface hardness of 2mm resin composite as compared to the hardness produced by the conventional halogen light $(400mW/cm^2)$.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.12C
• /
• pp.1288-1298
• /
• 2002

To implement elliptic curve cryptosystem in GF(2$\^$m/) at high speed, a fast divider is required. Although bit-parallel architecture is well suited for high speed division operations, elliptic curve cryptosystem requires large m(at least 163) to support a sufficient security. In other words, since the bit-parallel architecture has an area complexity of 0(m$\^$m/), it is not suited for this application. In this paper, we propose a new serial-in serial-out systolic array for computing division operations in GF(2$\^$m/) using the standard basis representation. Based on a modified version of tile binary extended greatest common divisor algorithm, we obtain a new data dependence graph and design an efficient bit-serial systolic divider. The proposed divider has 0(m) time complexity and 0(m) area complexity. If input data come in continuously, the proposed divider can produce division results at a rate of one per m clock cycles, after an initial delay of 5m-2 cycles. Analysis shows that the proposed divider provides a significant reduction in both chip area and computational delay time compared to previously proposed systolic dividers with the same I/O format. Since the proposed divider can perform division operations at high speed with the reduced chip area, it is well suited for division circuit of elliptic curve cryptosystem. Furthermore, since the proposed architecture does not restrict the choice of irreducible polynomial, and has a unidirectional data flow and regularity, it provides a high flexibility and scalability with respect to the field size m.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.20 no.2
• /
• pp.86-93
• /
• 2016

Performance tests of a 75-tonf liquid rocket engine turbopump were conducted. The performance of sub-components - two pumps and a turbine - and their power matching were measured and examined firstly near the design speed under the LN2 and kerosene environment. In the real propellant - LOX and kerosene - environment tests, design and off-design performance of turbopump were fully verified in regime of the rocket engine operation. During the off-design performance tests, turbopump running time was set longer than the engine operating time to verify the pump operability and set the pump inlet pressure close to design NPSHr to investigate pump suction capability in parallel. It has been found that developed-turbopump satisfied all of the engine required performances.

• Journal of Oral Medicine and Pain
• /
• v.31 no.3
• /
• pp.211-221
• /
• 2006

Purpose The aims of this study were to evaluate micro-tensile bond strength of composite resin bonded to dentin following high-speed rotary handpiece preparation or Er:YAG laser preparation with two different adhesive systems and to assess the influence of different Er:YAG laser energies on the micro-tensile bond strength. Materials and Methods In this study, 40 third morlars were used. Flat dentin specimans were obtained and randomly assigned to eight groups. Dentin surfaces were prepared with one of four cutting types: carbide bur, Er:YAG laser (2 W, 3 W and 4 W) and conditioned with two bonding systems, Scotchbond Multipurpose Plus (SM), Clearfil SE bond (SE) and composite resin-build ups were created. After storage for 24 hours, each specimen was serially sectioned perpendicular to the bonded surface to produce more than thirty slabs in each group. Micro-tensile bond strength test was performed at a crosshead speed of 1.0 mm/min. Micro-tensile bond strengths (${\mu}TBS$) were expressed as means$\pm$SD. Data were submitted to statistical analysis using two-way ANOVA, one-way ANOVA, Student-Newman-Keuls' multiple comparison test and t-test. Results and Conclusion 1. Regardless of bonding systems, the ${\mu}TBS$ according to cutting types were from highest to lowest : 3 W, 2 W, Bur, and 4 W. In addition, there was no significant difference between Bur and 4 W (p<0.001). 2. Regardless of cutting types, SM showed significantly higher ${\mu}TBS$ than SE (p<0.001). 3. Bonding to dentin conditioned with SM resulted in higher ${\mu}TBS$ for 3 W compared to Bur, 2 W, and 4 W. There was no significant difference between 2 W and Bur (p<0.001). 4. Bonding to dentin conditioned with SE resulted in higher ${\mu}TBS$ for 3 W compared to 2 W, 4 W, and Bur. Bur exhibited significant lower ${\mu}TBS$ than all other cutting types. There were no significant differences between 3 W, 2 W and between 4 W and Bur (p<0.001). 5. The ${\mu}TBS$ of laser cutting groups were shown in order from highest to lowest: 3 W, 2 W and 4 W in two bonding systems. There was no significant difference between 2 W and 3 W in SE (p<0.001). : The ${\mu}TBS$ of composite resin bonded dentin was significantly affected by interaction between the cutting type and bonding system. In the range of 2 W-3 W, cavity preparation of the Er:YAG laser seems to supply good adhesion of composite resin restoration no less than bur preparation. In particular, if you want to use the self-etching system, including Clearfil SE bond for the purpose of a simplification of the bonding procedures and prevention of adverse effects by excessive etching, an Er:YAG laser may offer better adhesion than a bur.

• Geophysics and Geophysical Exploration
• /
• v.8 no.3
• /
• pp.224-231
• /
• 2005

In geophysical field surveys, commercial equipments often fail to resolve the subsurface target or even sometimes fail to be applied because they do not fit to the various field situations or the physical properties of the medium or target. We developed a geophysical measurement system, which can be easily adapted for the various field situations and targets. The system based on PXI with A/D converter and some stand alone equipment such as Network Analyzer was applied to borehole radar survey, borehole sonic measurement and electromagnetic noise measurement. The system for borehole radar survey consists of PXI, Network Analyzer, dipole antennas, GPIB interface is used for PXI to control Network Analyzer. The system for borehole sonic measurement consists of PXI, 24 Bit A/D converter, high voltage pulse generator, transmitting and receiving piezoelectric sensors. The electromagnetic noise measurement system consists of PXI, 24 Bit A/D converter, 2 horizontal component electric field sensors and 2 horizontal and 1 vertical component magnetic filed sensors. The borehole radar system has been successfully applied to detect the width of the artificial tunnel through which the borehole pass and to image buried steel pipe, while the commercial borehole radar equipment failed. The borehole sonic system was tested to detect the width of artificial tunnel and showed a reasonable result. The characteristic of electromagnetic noise was grasped at an urban area with the data from the electromagnetic noise measurement system. The system is also applied to characterize the signal distortion by induction between the electric cables in resistivity survey. The system can be applied various geophysical problems with a simple modification of the system and sensors.

• Journal of IKEEE
• /
• v.25 no.3
• /
• pp.493-500
• /
• 2021

In this paper, we propose the development of deep learning structure to improve quality of polygonal containers. The deep learning structure consists of a convolution layer, a bottleneck layer, a fully connect layer, and a softmax layer. The convolution layer is a layer that obtains a feature image by performing a convolution 3x3 operation on the input image or the feature image of the previous layer with several feature filters. The bottleneck layer selects only the optimal features among the features on the feature image extracted through the convolution layer, reduces the channel to a convolution 1x1 ReLU, and performs a convolution 3x3 ReLU. The global average pooling operation performed after going through the bottleneck layer reduces the size of the feature image by selecting only the optimal features among the features of the feature image extracted through the convolution layer. The fully connect layer outputs the output data through 6 fully connect layers. The softmax layer multiplies and multiplies the value between the value of the input layer node and the target node to be calculated, and converts it into a value between 0 and 1 through an activation function. After the learning is completed, the recognition process classifies non-circular glass bottles by performing image acquisition using a camera, measuring position detection, and non-circular glass bottle classification using deep learning as in the learning process. In order to evaluate the performance of the deep learning structure to improve quality of polygonal containers, as a result of an experiment at an authorized testing institute, it was calculated to be at the same level as the world's highest level with 99% good/defective discrimination accuracy. Inspection time averaged 1.7 seconds, which was calculated within the operating time standards of production processes using non-circular machine vision systems. Therefore, the effectiveness of the performance of the deep learning structure to improve quality of polygonal containers proposed in this paper was proven.

• Korean Journal of Remote Sensing
• /
• v.39 no.2
• /
• pp.207-221
• /
• 2023

This study suggests deep neural network models for estimating air temperature with Level 1B (L1B) datasets of GEO-KOMPSAT-2A (GK-2A). The temperature at 1.5 m above the ground impact not only daily life but also weather warnings such as cold and heat waves. There are many studies to assume the air temperature from the land surface temperature (LST) retrieved from satellites because the air temperature has a strong relationship with the LST. However, an algorithm of the LST, Level 2 output of GK-2A, works only clear sky pixels. To overcome the cloud effects, we apply a deep neural network (DNN) model to assume the air temperature with L1B calibrated for radiometric and geometrics from raw satellite data and compare the model with a linear regression model between LST and air temperature. The root mean square errors (RMSE) of the air temperature for model outputs are used to evaluate the model. The number of 95 in-situ air temperature data was 2,496,634 and the ratio of datasets paired with LST and L1B show 42.1% and 98.4%. The training years are 2020 and 2021 and 2022 is used to validate. The DNN model is designed with an input layer taking 16 channels and four hidden fully connected layers to assume an air temperature. As a result of the model using 16 bands of L1B, the DNN with RMSE 2.22℃ showed great performance than the baseline model with RMSE 3.55℃ on clear sky conditions and the total RMSE including overcast samples was 3.33℃. It is suggested that the DNN is able to overcome cloud effects. However, it showed different characteristics in seasonal and hourly analysis and needed to append solar information as inputs to make a general DNN model because the summer and winter seasons showed a low coefficient of determinations with high standard deviations.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• Radiation Oncology Journal
• /
• v.10 no.1
• /
• pp.7-14
• /
• 1992

Between August 1988 and December 1991, 24 patients with intracranial tumors were treated with stereotactic radiosurgery(RS) using a 10 MV linear accelerator at Severance Hospital, Yonsei University College of Medicine. There were 5 meningiomas, 3 craniopharyngiomas, 9 glial tumors, 2 solitary metastases, 2 acoustic neurinomas, 2 pineal tumors, and 1 non-Hodgkin's lymphoma. Ten patients were treated as primary treatment after diagnosis with stereotactic biopsy or neuroimaging study. Nine patients underwent RS for post-op. residual tumors and three patients as a salvage treatment for recurrence after external irradiation. Two patients received RS as a boost followed by fractionated conventional radiotherapy. Among sixteen patients who were followed more than 6 months with neuroimage, seven patients (2 meningiomas, 4 benign glial tumors, one non-Hodgkin's lymphoma) showed complete response on neuroimage after RS and nine patients showed decreased tumor size. There was no acute treatment related side reaction. Late complications include three patients with symptomatic peritumoral brain edema and one craniopharyngioma with optic chiasmal injury. Through this early experience, we conclude that stereotactically directed single high doses of irradiation to the small intracranial tumors is effective for tumor control. However, in order to define the role of radiosurgery in the management of intracraniai tumors, we should get the long-term results available to demonstrate the benefits versus potential complications of this therapeutic modality.